Hugging Face's logo

pharaouk
/
loki2

Model card Files Community
Farouk commited on
Commit
13d8fa3
·
1 Parent(s): a544699

Training in progress, step 600

Browse files
Files changed (16) hide show
  1. adapter_model.bin +1 -1
  2. checkpoint-400/adapter_model/adapter_model/README.md +20 -0
  3. checkpoint-400/adapter_model/adapter_model/adapter_config.json +26 -0
  4. checkpoint-400/adapter_model/adapter_model/adapter_model.bin +3 -0
  5. checkpoint-600/README.md +20 -0
  6. checkpoint-600/adapter_config.json +26 -0
  7. checkpoint-600/adapter_model.bin +3 -0
  8. checkpoint-600/added_tokens.json +3 -0
  9. checkpoint-600/optimizer.pt +3 -0
  10. checkpoint-600/rng_state.pth +3 -0
  11. checkpoint-600/scheduler.pt +3 -0
  12. checkpoint-600/special_tokens_map.json +6 -0
  13. checkpoint-600/tokenizer.model +3 -0
  14. checkpoint-600/tokenizer_config.json +35 -0
  15. checkpoint-600/trainer_state.json +4168 -0
  16. checkpoint-600/training_args.bin +3 -0
adapter_model.bin CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:8b42360731aeed8d0662eff0fa1259bd55a45091b5078ae386cbce2d3325842b
3
  size 319977229
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:83cb2b2d2fd357436ba4639eca9270b4833144a38d675fa83baa39e0efd9500f
3
  size 319977229
checkpoint-400/adapter_model/adapter_model/README.md ADDED
@@ -0,0 +1,20 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ ---
2
+ library_name: peft
3
+ ---
4
+ ## Training procedure
5
+
6
+
7
+ The following `bitsandbytes` quantization config was used during training:
8
+ - load_in_8bit: False
9
+ - load_in_4bit: True
10
+ - llm_int8_threshold: 6.0
11
+ - llm_int8_skip_modules: None
12
+ - llm_int8_enable_fp32_cpu_offload: False
13
+ - llm_int8_has_fp16_weight: False
14
+ - bnb_4bit_quant_type: nf4
15
+ - bnb_4bit_use_double_quant: True
16
+ - bnb_4bit_compute_dtype: bfloat16
17
+ ### Framework versions
18
+
19
+
20
+ - PEFT 0.4.0
checkpoint-400/adapter_model/adapter_model/adapter_config.json ADDED
@@ -0,0 +1,26 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "auto_mapping": null,
3
+ "base_model_name_or_path": "NousResearch/Nous-Hermes-llama-2-7b",
4
+ "bias": "none",
5
+ "fan_in_fan_out": false,
6
+ "inference_mode": true,
7
+ "init_lora_weights": true,
8
+ "layers_pattern": null,
9
+ "layers_to_transform": null,
10
+ "lora_alpha": 16.0,
11
+ "lora_dropout": 0.4,
12
+ "modules_to_save": null,
13
+ "peft_type": "LORA",
14
+ "r": 64,
15
+ "revision": null,
16
+ "target_modules": [
17
+ "o_proj",
18
+ "v_proj",
19
+ "k_proj",
20
+ "down_proj",
21
+ "q_proj",
22
+ "up_proj",
23
+ "gate_proj"
24
+ ],
25
+ "task_type": "CAUSAL_LM"
26
+ }
checkpoint-400/adapter_model/adapter_model/adapter_model.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:8b42360731aeed8d0662eff0fa1259bd55a45091b5078ae386cbce2d3325842b
3
+ size 319977229
checkpoint-600/README.md ADDED
@@ -0,0 +1,20 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ ---
2
+ library_name: peft
3
+ ---
4
+ ## Training procedure
5
+
6
+
7
+ The following `bitsandbytes` quantization config was used during training:
8
+ - load_in_8bit: False
9
+ - load_in_4bit: True
10
+ - llm_int8_threshold: 6.0
11
+ - llm_int8_skip_modules: None
12
+ - llm_int8_enable_fp32_cpu_offload: False
13
+ - llm_int8_has_fp16_weight: False
14
+ - bnb_4bit_quant_type: nf4
15
+ - bnb_4bit_use_double_quant: True
16
+ - bnb_4bit_compute_dtype: bfloat16
17
+ ### Framework versions
18
+
19
+
20
+ - PEFT 0.4.0
checkpoint-600/adapter_config.json ADDED
@@ -0,0 +1,26 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "auto_mapping": null,
3
+ "base_model_name_or_path": "NousResearch/Nous-Hermes-llama-2-7b",
4
+ "bias": "none",
5
+ "fan_in_fan_out": false,
6
+ "inference_mode": true,
7
+ "init_lora_weights": true,
8
+ "layers_pattern": null,
9
+ "layers_to_transform": null,
10
+ "lora_alpha": 16.0,
11
+ "lora_dropout": 0.4,
12
+ "modules_to_save": null,
13
+ "peft_type": "LORA",
14
+ "r": 64,
15
+ "revision": null,
16
+ "target_modules": [
17
+ "o_proj",
18
+ "v_proj",
19
+ "k_proj",
20
+ "down_proj",
21
+ "q_proj",
22
+ "up_proj",
23
+ "gate_proj"
24
+ ],
25
+ "task_type": "CAUSAL_LM"
26
+ }
checkpoint-600/adapter_model.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:83cb2b2d2fd357436ba4639eca9270b4833144a38d675fa83baa39e0efd9500f
3
+ size 319977229
checkpoint-600/added_tokens.json ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ {
2
+ "<pad>": 32000
3
+ }
checkpoint-600/optimizer.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:1814ba7f35163ffaf78d345d4b39c7a8b3ed8f496cae368dcc88b0f4925a89cf
3
+ size 1279539973
checkpoint-600/rng_state.pth ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:428800511b314d28f9b52a9e7f3e52507c9b3807682a1f0b719700359060bcc4
3
+ size 14511
checkpoint-600/scheduler.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:ec86fe6259605d8ef886cd8fc711f042f8248ddd86839f789c7a9c4f568fa3a5
3
+ size 627
checkpoint-600/special_tokens_map.json ADDED
@@ -0,0 +1,6 @@
 
 
 
 
 
 
 
1
+ {
2
+ "bos_token": "<s>",
3
+ "eos_token": "</s>",
4
+ "pad_token": "<unk>",
5
+ "unk_token": "<unk>"
6
+ }
checkpoint-600/tokenizer.model ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:9e556afd44213b6bd1be2b850ebbbd98f5481437a8021afaf58ee7fb1818d347
3
+ size 499723
checkpoint-600/tokenizer_config.json ADDED
@@ -0,0 +1,35 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "add_bos_token": true,
3
+ "add_eos_token": false,
4
+ "bos_token": {
5
+ "__type": "AddedToken",
6
+ "content": "<s>",
7
+ "lstrip": false,
8
+ "normalized": true,
9
+ "rstrip": false,
10
+ "single_word": false
11
+ },
12
+ "clean_up_tokenization_spaces": false,
13
+ "eos_token": {
14
+ "__type": "AddedToken",
15
+ "content": "</s>",
16
+ "lstrip": false,
17
+ "normalized": true,
18
+ "rstrip": false,
19
+ "single_word": false
20
+ },
21
+ "legacy": false,
22
+ "model_max_length": 1000000000000000019884624838656,
23
+ "pad_token": null,
24
+ "padding_side": "right",
25
+ "sp_model_kwargs": {},
26
+ "tokenizer_class": "LlamaTokenizer",
27
+ "unk_token": {
28
+ "__type": "AddedToken",
29
+ "content": "<unk>",
30
+ "lstrip": false,
31
+ "normalized": true,
32
+ "rstrip": false,
33
+ "single_word": false
34
+ }
35
+ }
checkpoint-600/trainer_state.json ADDED
@@ -0,0 +1,4168 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "best_metric": 0.5055681467056274,
3
+ "best_model_checkpoint": "./output_v2/7b_cluster00_Nous-Hermes-llama-2-7b_codellama_blob_1/checkpoint-600",
4
+ "epoch": 1.6842105263157894,
5
+ "global_step": 600,
6
+ "is_hyper_param_search": false,
7
+ "is_local_process_zero": true,
8
+ "is_world_process_zero": true,
9
+ "log_history": [
10
+ {
11
+ "epoch": 0.0,
12
+ "learning_rate": 9.999999999999999e-06,
13
+ "loss": 0.7795,
14
+ "step": 1
15
+ },
16
+ {
17
+ "epoch": 0.01,
18
+ "learning_rate": 1.9999999999999998e-05,
19
+ "loss": 0.6789,
20
+ "step": 2
21
+ },
22
+ {
23
+ "epoch": 0.01,
24
+ "learning_rate": 2.9999999999999997e-05,
25
+ "loss": 0.6846,
26
+ "step": 3
27
+ },
28
+ {
29
+ "epoch": 0.01,
30
+ "learning_rate": 3.9999999999999996e-05,
31
+ "loss": 0.6883,
32
+ "step": 4
33
+ },
34
+ {
35
+ "epoch": 0.01,
36
+ "learning_rate": 4.9999999999999996e-05,
37
+ "loss": 0.5963,
38
+ "step": 5
39
+ },
40
+ {
41
+ "epoch": 0.02,
42
+ "learning_rate": 5.9999999999999995e-05,
43
+ "loss": 0.6577,
44
+ "step": 6
45
+ },
46
+ {
47
+ "epoch": 0.02,
48
+ "learning_rate": 7e-05,
49
+ "loss": 0.625,
50
+ "step": 7
51
+ },
52
+ {
53
+ "epoch": 0.02,
54
+ "learning_rate": 7.999999999999999e-05,
55
+ "loss": 0.6864,
56
+ "step": 8
57
+ },
58
+ {
59
+ "epoch": 0.03,
60
+ "learning_rate": 8.999999999999999e-05,
61
+ "loss": 0.5344,
62
+ "step": 9
63
+ },
64
+ {
65
+ "epoch": 0.03,
66
+ "learning_rate": 9.999999999999999e-05,
67
+ "loss": 0.5956,
68
+ "step": 10
69
+ },
70
+ {
71
+ "epoch": 0.03,
72
+ "learning_rate": 0.00010999999999999998,
73
+ "loss": 0.6993,
74
+ "step": 11
75
+ },
76
+ {
77
+ "epoch": 0.03,
78
+ "learning_rate": 0.00011999999999999999,
79
+ "loss": 0.623,
80
+ "step": 12
81
+ },
82
+ {
83
+ "epoch": 0.04,
84
+ "learning_rate": 0.00013,
85
+ "loss": 0.5547,
86
+ "step": 13
87
+ },
88
+ {
89
+ "epoch": 0.04,
90
+ "learning_rate": 0.00014,
91
+ "loss": 0.5404,
92
+ "step": 14
93
+ },
94
+ {
95
+ "epoch": 0.04,
96
+ "learning_rate": 0.00015,
97
+ "loss": 0.5536,
98
+ "step": 15
99
+ },
100
+ {
101
+ "epoch": 0.04,
102
+ "learning_rate": 0.00015999999999999999,
103
+ "loss": 0.6886,
104
+ "step": 16
105
+ },
106
+ {
107
+ "epoch": 0.05,
108
+ "learning_rate": 0.00016999999999999999,
109
+ "loss": 0.8917,
110
+ "step": 17
111
+ },
112
+ {
113
+ "epoch": 0.05,
114
+ "learning_rate": 0.00017999999999999998,
115
+ "loss": 0.6834,
116
+ "step": 18
117
+ },
118
+ {
119
+ "epoch": 0.05,
120
+ "learning_rate": 0.00018999999999999998,
121
+ "loss": 0.4784,
122
+ "step": 19
123
+ },
124
+ {
125
+ "epoch": 0.06,
126
+ "learning_rate": 0.00019999999999999998,
127
+ "loss": 0.5869,
128
+ "step": 20
129
+ },
130
+ {
131
+ "epoch": 0.06,
132
+ "learning_rate": 0.00020999999999999998,
133
+ "loss": 0.9495,
134
+ "step": 21
135
+ },
136
+ {
137
+ "epoch": 0.06,
138
+ "learning_rate": 0.00021999999999999995,
139
+ "loss": 0.8727,
140
+ "step": 22
141
+ },
142
+ {
143
+ "epoch": 0.06,
144
+ "learning_rate": 0.00023,
145
+ "loss": 0.671,
146
+ "step": 23
147
+ },
148
+ {
149
+ "epoch": 0.07,
150
+ "learning_rate": 0.00023999999999999998,
151
+ "loss": 0.6582,
152
+ "step": 24
153
+ },
154
+ {
155
+ "epoch": 0.07,
156
+ "learning_rate": 0.00025,
157
+ "loss": 0.6599,
158
+ "step": 25
159
+ },
160
+ {
161
+ "epoch": 0.07,
162
+ "eval_loss": 0.6080610752105713,
163
+ "eval_runtime": 26.3081,
164
+ "eval_samples_per_second": 7.602,
165
+ "eval_steps_per_second": 1.901,
166
+ "step": 25
167
+ },
168
+ {
169
+ "dharma_eval_accuracy": 0.45126789906203946,
170
+ "dharma_eval_accuracy_ARC-Challenge": 0.5555555555555556,
171
+ "dharma_eval_accuracy_ARC-Easy": 0.5555555555555556,
172
+ "dharma_eval_accuracy_BoolQ": 0.7037037037037037,
173
+ "dharma_eval_accuracy_MMLU": 0.3888888888888889,
174
+ "dharma_eval_accuracy_agieval": 0.2711864406779661,
175
+ "dharma_eval_accuracy_bigbench": 0.47540983606557374,
176
+ "dharma_eval_accuracy_openbookqa": 0.35185185185185186,
177
+ "dharma_eval_accuracy_truthful_qa": 0.42592592592592593,
178
+ "dharma_eval_accuracy_winogrande": 0.3333333333333333,
179
+ "dharma_loss": 2.3038411536216734,
180
+ "epoch": 0.07,
181
+ "step": 25
182
+ },
183
+ {
184
+ "epoch": 0.07,
185
+ "learning_rate": 0.00026,
186
+ "loss": 0.568,
187
+ "step": 26
188
+ },
189
+ {
190
+ "epoch": 0.08,
191
+ "learning_rate": 0.00027,
192
+ "loss": 0.569,
193
+ "step": 27
194
+ },
195
+ {
196
+ "epoch": 0.08,
197
+ "learning_rate": 0.00028,
198
+ "loss": 0.4481,
199
+ "step": 28
200
+ },
201
+ {
202
+ "epoch": 0.08,
203
+ "learning_rate": 0.00029,
204
+ "loss": 0.5482,
205
+ "step": 29
206
+ },
207
+ {
208
+ "epoch": 0.08,
209
+ "learning_rate": 0.0003,
210
+ "loss": 0.8189,
211
+ "step": 30
212
+ },
213
+ {
214
+ "epoch": 0.09,
215
+ "learning_rate": 0.0002999992132854894,
216
+ "loss": 0.928,
217
+ "step": 31
218
+ },
219
+ {
220
+ "epoch": 0.09,
221
+ "learning_rate": 0.0002999968531502098,
222
+ "loss": 0.6567,
223
+ "step": 32
224
+ },
225
+ {
226
+ "epoch": 0.09,
227
+ "learning_rate": 0.0002999929196189179,
228
+ "loss": 0.469,
229
+ "step": 33
230
+ },
231
+ {
232
+ "epoch": 0.1,
233
+ "learning_rate": 0.00029998741273287477,
234
+ "loss": 0.6043,
235
+ "step": 34
236
+ },
237
+ {
238
+ "epoch": 0.1,
239
+ "learning_rate": 0.0002999803325498448,
240
+ "loss": 0.6288,
241
+ "step": 35
242
+ },
243
+ {
244
+ "epoch": 0.1,
245
+ "learning_rate": 0.0002999716791440959,
246
+ "loss": 0.6692,
247
+ "step": 36
248
+ },
249
+ {
250
+ "epoch": 0.1,
251
+ "learning_rate": 0.00029996145260639806,
252
+ "loss": 0.7591,
253
+ "step": 37
254
+ },
255
+ {
256
+ "epoch": 0.11,
257
+ "learning_rate": 0.000299949653044023,
258
+ "loss": 0.5729,
259
+ "step": 38
260
+ },
261
+ {
262
+ "epoch": 0.11,
263
+ "learning_rate": 0.00029993628058074245,
264
+ "loss": 0.4347,
265
+ "step": 39
266
+ },
267
+ {
268
+ "epoch": 0.11,
269
+ "learning_rate": 0.00029992133535682725,
270
+ "loss": 0.7718,
271
+ "step": 40
272
+ },
273
+ {
274
+ "epoch": 0.12,
275
+ "learning_rate": 0.00029990481752904563,
276
+ "loss": 0.7768,
277
+ "step": 41
278
+ },
279
+ {
280
+ "epoch": 0.12,
281
+ "learning_rate": 0.0002998867272706619,
282
+ "loss": 0.7359,
283
+ "step": 42
284
+ },
285
+ {
286
+ "epoch": 0.12,
287
+ "learning_rate": 0.0002998670647714343,
288
+ "loss": 0.5626,
289
+ "step": 43
290
+ },
291
+ {
292
+ "epoch": 0.12,
293
+ "learning_rate": 0.00029984583023761317,
294
+ "loss": 0.5444,
295
+ "step": 44
296
+ },
297
+ {
298
+ "epoch": 0.13,
299
+ "learning_rate": 0.00029982302389193856,
300
+ "loss": 0.5487,
301
+ "step": 45
302
+ },
303
+ {
304
+ "epoch": 0.13,
305
+ "learning_rate": 0.00029979864597363845,
306
+ "loss": 0.4916,
307
+ "step": 46
308
+ },
309
+ {
310
+ "epoch": 0.13,
311
+ "learning_rate": 0.0002997726967384255,
312
+ "loss": 0.4817,
313
+ "step": 47
314
+ },
315
+ {
316
+ "epoch": 0.13,
317
+ "learning_rate": 0.00029974517645849503,
318
+ "loss": 0.6375,
319
+ "step": 48
320
+ },
321
+ {
322
+ "epoch": 0.14,
323
+ "learning_rate": 0.0002997160854225217,
324
+ "loss": 0.5938,
325
+ "step": 49
326
+ },
327
+ {
328
+ "epoch": 0.14,
329
+ "learning_rate": 0.0002996854239356567,
330
+ "loss": 0.5868,
331
+ "step": 50
332
+ },
333
+ {
334
+ "epoch": 0.14,
335
+ "eval_loss": 0.5586252212524414,
336
+ "eval_runtime": 26.2488,
337
+ "eval_samples_per_second": 7.619,
338
+ "eval_steps_per_second": 1.905,
339
+ "step": 50
340
+ },
341
+ {
342
+ "dharma_eval_accuracy": 0.4279315127544574,
343
+ "dharma_eval_accuracy_ARC-Challenge": 0.5925925925925926,
344
+ "dharma_eval_accuracy_ARC-Easy": 0.6481481481481481,
345
+ "dharma_eval_accuracy_BoolQ": 0.7037037037037037,
346
+ "dharma_eval_accuracy_MMLU": 0.4074074074074074,
347
+ "dharma_eval_accuracy_agieval": 0.2542372881355932,
348
+ "dharma_eval_accuracy_bigbench": 0.39344262295081966,
349
+ "dharma_eval_accuracy_openbookqa": 0.3148148148148148,
350
+ "dharma_eval_accuracy_truthful_qa": 0.4444444444444444,
351
+ "dharma_eval_accuracy_winogrande": 0.09259259259259259,
352
+ "dharma_loss": 2.1031297972202303,
353
+ "epoch": 0.14,
354
+ "step": 50
355
+ },
356
+ {
357
+ "epoch": 0.14,
358
+ "learning_rate": 0.00029965319231952455,
359
+ "loss": 0.612,
360
+ "step": 51
361
+ },
362
+ {
363
+ "epoch": 0.15,
364
+ "learning_rate": 0.0002996193909122197,
365
+ "loss": 0.8942,
366
+ "step": 52
367
+ },
368
+ {
369
+ "epoch": 0.15,
370
+ "learning_rate": 0.00029958402006830274,
371
+ "loss": 0.5766,
372
+ "step": 53
373
+ },
374
+ {
375
+ "epoch": 0.15,
376
+ "learning_rate": 0.0002995470801587973,
377
+ "loss": 0.6238,
378
+ "step": 54
379
+ },
380
+ {
381
+ "epoch": 0.15,
382
+ "learning_rate": 0.0002995085715711854,
383
+ "loss": 0.5561,
384
+ "step": 55
385
+ },
386
+ {
387
+ "epoch": 0.16,
388
+ "learning_rate": 0.0002994684947094039,
389
+ "loss": 0.6203,
390
+ "step": 56
391
+ },
392
+ {
393
+ "epoch": 0.16,
394
+ "learning_rate": 0.0002994268499938403,
395
+ "loss": 0.6534,
396
+ "step": 57
397
+ },
398
+ {
399
+ "epoch": 0.16,
400
+ "learning_rate": 0.00029938363786132774,
401
+ "loss": 0.594,
402
+ "step": 58
403
+ },
404
+ {
405
+ "epoch": 0.17,
406
+ "learning_rate": 0.00029933885876514115,
407
+ "loss": 0.5608,
408
+ "step": 59
409
+ },
410
+ {
411
+ "epoch": 0.17,
412
+ "learning_rate": 0.0002992925131749921,
413
+ "loss": 0.6622,
414
+ "step": 60
415
+ },
416
+ {
417
+ "epoch": 0.17,
418
+ "learning_rate": 0.00029924460157702376,
419
+ "loss": 0.5207,
420
+ "step": 61
421
+ },
422
+ {
423
+ "epoch": 0.17,
424
+ "learning_rate": 0.00029919512447380625,
425
+ "loss": 0.5308,
426
+ "step": 62
427
+ },
428
+ {
429
+ "epoch": 0.18,
430
+ "learning_rate": 0.0002991440823843309,
431
+ "loss": 0.6217,
432
+ "step": 63
433
+ },
434
+ {
435
+ "epoch": 0.18,
436
+ "learning_rate": 0.0002990914758440051,
437
+ "loss": 0.6541,
438
+ "step": 64
439
+ },
440
+ {
441
+ "epoch": 0.18,
442
+ "learning_rate": 0.00029903730540464666,
443
+ "loss": 0.514,
444
+ "step": 65
445
+ },
446
+ {
447
+ "epoch": 0.19,
448
+ "learning_rate": 0.00029898157163447763,
449
+ "loss": 0.5923,
450
+ "step": 66
451
+ },
452
+ {
453
+ "epoch": 0.19,
454
+ "learning_rate": 0.0002989242751181191,
455
+ "loss": 0.6156,
456
+ "step": 67
457
+ },
458
+ {
459
+ "epoch": 0.19,
460
+ "learning_rate": 0.0002988654164565843,
461
+ "loss": 0.5406,
462
+ "step": 68
463
+ },
464
+ {
465
+ "epoch": 0.19,
466
+ "learning_rate": 0.0002988049962672728,
467
+ "loss": 0.4563,
468
+ "step": 69
469
+ },
470
+ {
471
+ "epoch": 0.2,
472
+ "learning_rate": 0.00029874301518396376,
473
+ "loss": 0.5448,
474
+ "step": 70
475
+ },
476
+ {
477
+ "epoch": 0.2,
478
+ "learning_rate": 0.00029867947385680936,
479
+ "loss": 0.6164,
480
+ "step": 71
481
+ },
482
+ {
483
+ "epoch": 0.2,
484
+ "learning_rate": 0.0002986143729523282,
485
+ "loss": 0.5636,
486
+ "step": 72
487
+ },
488
+ {
489
+ "epoch": 0.2,
490
+ "learning_rate": 0.00029854771315339785,
491
+ "loss": 0.5301,
492
+ "step": 73
493
+ },
494
+ {
495
+ "epoch": 0.21,
496
+ "learning_rate": 0.00029847949515924806,
497
+ "loss": 0.6074,
498
+ "step": 74
499
+ },
500
+ {
501
+ "epoch": 0.21,
502
+ "learning_rate": 0.0002984097196854534,
503
+ "loss": 0.5123,
504
+ "step": 75
505
+ },
506
+ {
507
+ "epoch": 0.21,
508
+ "eval_loss": 0.541282057762146,
509
+ "eval_runtime": 26.1496,
510
+ "eval_samples_per_second": 7.648,
511
+ "eval_steps_per_second": 1.912,
512
+ "step": 75
513
+ },
514
+ {
515
+ "dharma_eval_accuracy": 0.4466804336366869,
516
+ "dharma_eval_accuracy_ARC-Challenge": 0.6296296296296297,
517
+ "dharma_eval_accuracy_ARC-Easy": 0.5740740740740741,
518
+ "dharma_eval_accuracy_BoolQ": 0.7037037037037037,
519
+ "dharma_eval_accuracy_MMLU": 0.4074074074074074,
520
+ "dharma_eval_accuracy_agieval": 0.2711864406779661,
521
+ "dharma_eval_accuracy_bigbench": 0.5081967213114754,
522
+ "dharma_eval_accuracy_openbookqa": 0.12962962962962962,
523
+ "dharma_eval_accuracy_truthful_qa": 0.4074074074074074,
524
+ "dharma_eval_accuracy_winogrande": 0.3888888888888889,
525
+ "dharma_loss": 2.2732545466423035,
526
+ "epoch": 0.21,
527
+ "step": 75
528
+ },
529
+ {
530
+ "epoch": 0.21,
531
+ "learning_rate": 0.0002983383874639254,
532
+ "loss": 0.5206,
533
+ "step": 76
534
+ },
535
+ {
536
+ "epoch": 0.22,
537
+ "learning_rate": 0.00029826549924290557,
538
+ "loss": 0.6452,
539
+ "step": 77
540
+ },
541
+ {
542
+ "epoch": 0.22,
543
+ "learning_rate": 0.00029819105578695655,
544
+ "loss": 0.6085,
545
+ "step": 78
546
+ },
547
+ {
548
+ "epoch": 0.22,
549
+ "learning_rate": 0.00029811505787695524,
550
+ "loss": 0.5163,
551
+ "step": 79
552
+ },
553
+ {
554
+ "epoch": 0.22,
555
+ "learning_rate": 0.00029803750631008356,
556
+ "loss": 0.5511,
557
+ "step": 80
558
+ },
559
+ {
560
+ "epoch": 0.23,
561
+ "learning_rate": 0.0002979584018998209,
562
+ "loss": 0.6395,
563
+ "step": 81
564
+ },
565
+ {
566
+ "epoch": 0.23,
567
+ "learning_rate": 0.000297877745475935,
568
+ "loss": 0.6281,
569
+ "step": 82
570
+ },
571
+ {
572
+ "epoch": 0.23,
573
+ "learning_rate": 0.00029779553788447357,
574
+ "loss": 0.4932,
575
+ "step": 83
576
+ },
577
+ {
578
+ "epoch": 0.24,
579
+ "learning_rate": 0.00029771177998775536,
580
+ "loss": 0.5497,
581
+ "step": 84
582
+ },
583
+ {
584
+ "epoch": 0.24,
585
+ "learning_rate": 0.0002976264726643611,
586
+ "loss": 0.5496,
587
+ "step": 85
588
+ },
589
+ {
590
+ "epoch": 0.24,
591
+ "learning_rate": 0.00029753961680912427,
592
+ "loss": 0.4569,
593
+ "step": 86
594
+ },
595
+ {
596
+ "epoch": 0.24,
597
+ "learning_rate": 0.00029745121333312163,
598
+ "loss": 0.5152,
599
+ "step": 87
600
+ },
601
+ {
602
+ "epoch": 0.25,
603
+ "learning_rate": 0.0002973612631636638,
604
+ "loss": 0.6928,
605
+ "step": 88
606
+ },
607
+ {
608
+ "epoch": 0.25,
609
+ "learning_rate": 0.0002972697672442856,
610
+ "loss": 0.5503,
611
+ "step": 89
612
+ },
613
+ {
614
+ "epoch": 0.25,
615
+ "learning_rate": 0.0002971767265347358,
616
+ "loss": 0.7074,
617
+ "step": 90
618
+ },
619
+ {
620
+ "epoch": 0.26,
621
+ "learning_rate": 0.00029708214201096755,
622
+ "loss": 0.6397,
623
+ "step": 91
624
+ },
625
+ {
626
+ "epoch": 0.26,
627
+ "learning_rate": 0.0002969860146651276,
628
+ "loss": 0.7815,
629
+ "step": 92
630
+ },
631
+ {
632
+ "epoch": 0.26,
633
+ "learning_rate": 0.00029688834550554646,
634
+ "loss": 0.6274,
635
+ "step": 93
636
+ },
637
+ {
638
+ "epoch": 0.26,
639
+ "learning_rate": 0.0002967891355567273,
640
+ "loss": 0.6196,
641
+ "step": 94
642
+ },
643
+ {
644
+ "epoch": 0.27,
645
+ "learning_rate": 0.00029668838585933556,
646
+ "loss": 0.5845,
647
+ "step": 95
648
+ },
649
+ {
650
+ "epoch": 0.27,
651
+ "learning_rate": 0.0002965860974701879,
652
+ "loss": 0.5934,
653
+ "step": 96
654
+ },
655
+ {
656
+ "epoch": 0.27,
657
+ "learning_rate": 0.0002964822714622412,
658
+ "loss": 0.6277,
659
+ "step": 97
660
+ },
661
+ {
662
+ "epoch": 0.28,
663
+ "learning_rate": 0.000296376908924581,
664
+ "loss": 0.555,
665
+ "step": 98
666
+ },
667
+ {
668
+ "epoch": 0.28,
669
+ "learning_rate": 0.00029627001096241057,
670
+ "loss": 0.5478,
671
+ "step": 99
672
+ },
673
+ {
674
+ "epoch": 0.28,
675
+ "learning_rate": 0.0002961615786970389,
676
+ "loss": 0.6851,
677
+ "step": 100
678
+ },
679
+ {
680
+ "epoch": 0.28,
681
+ "eval_loss": 0.5322936177253723,
682
+ "eval_runtime": 26.2408,
683
+ "eval_samples_per_second": 7.622,
684
+ "eval_steps_per_second": 1.905,
685
+ "step": 100
686
+ },
687
+ {
688
+ "dharma_eval_accuracy": 0.3837211296690781,
689
+ "dharma_eval_accuracy_ARC-Challenge": 0.6296296296296297,
690
+ "dharma_eval_accuracy_ARC-Easy": 0.5925925925925926,
691
+ "dharma_eval_accuracy_BoolQ": 0.7037037037037037,
692
+ "dharma_eval_accuracy_MMLU": 0.4444444444444444,
693
+ "dharma_eval_accuracy_agieval": 0.22033898305084745,
694
+ "dharma_eval_accuracy_bigbench": 0.3442622950819672,
695
+ "dharma_eval_accuracy_openbookqa": 0.09259259259259259,
696
+ "dharma_eval_accuracy_truthful_qa": 0.2777777777777778,
697
+ "dharma_eval_accuracy_winogrande": 0.14814814814814814,
698
+ "dharma_loss": 2.1400452837944033,
699
+ "epoch": 0.28,
700
+ "step": 100
701
+ },
702
+ {
703
+ "epoch": 0.28,
704
+ "learning_rate": 0.00029605161326586916,
705
+ "loss": 0.5637,
706
+ "step": 101
707
+ },
708
+ {
709
+ "epoch": 0.29,
710
+ "learning_rate": 0.0002959401158223867,
711
+ "loss": 0.5563,
712
+ "step": 102
713
+ },
714
+ {
715
+ "epoch": 0.29,
716
+ "learning_rate": 0.0002958270875361469,
717
+ "loss": 0.7583,
718
+ "step": 103
719
+ },
720
+ {
721
+ "epoch": 0.29,
722
+ "learning_rate": 0.0002957125295927631,
723
+ "loss": 0.4775,
724
+ "step": 104
725
+ },
726
+ {
727
+ "epoch": 0.29,
728
+ "learning_rate": 0.0002955964431938939,
729
+ "loss": 0.6912,
730
+ "step": 105
731
+ },
732
+ {
733
+ "epoch": 0.3,
734
+ "learning_rate": 0.0002954788295572305,
735
+ "loss": 0.7142,
736
+ "step": 106
737
+ },
738
+ {
739
+ "epoch": 0.3,
740
+ "learning_rate": 0.00029535968991648454,
741
+ "loss": 0.5264,
742
+ "step": 107
743
+ },
744
+ {
745
+ "epoch": 0.3,
746
+ "learning_rate": 0.00029523902552137433,
747
+ "loss": 0.509,
748
+ "step": 108
749
+ },
750
+ {
751
+ "epoch": 0.31,
752
+ "learning_rate": 0.0002951168376376124,
753
+ "loss": 0.7375,
754
+ "step": 109
755
+ },
756
+ {
757
+ "epoch": 0.31,
758
+ "learning_rate": 0.0002949931275468917,
759
+ "loss": 0.4407,
760
+ "step": 110
761
+ },
762
+ {
763
+ "epoch": 0.31,
764
+ "learning_rate": 0.00029486789654687253,
765
+ "loss": 0.5361,
766
+ "step": 111
767
+ },
768
+ {
769
+ "epoch": 0.31,
770
+ "learning_rate": 0.00029474114595116896,
771
+ "loss": 0.6161,
772
+ "step": 112
773
+ },
774
+ {
775
+ "epoch": 0.32,
776
+ "learning_rate": 0.00029461287708933473,
777
+ "loss": 0.589,
778
+ "step": 113
779
+ },
780
+ {
781
+ "epoch": 0.32,
782
+ "learning_rate": 0.0002944830913068494,
783
+ "loss": 0.5677,
784
+ "step": 114
785
+ },
786
+ {
787
+ "epoch": 0.32,
788
+ "learning_rate": 0.00029435178996510455,
789
+ "loss": 0.5946,
790
+ "step": 115
791
+ },
792
+ {
793
+ "epoch": 0.33,
794
+ "learning_rate": 0.00029421897444138897,
795
+ "loss": 0.526,
796
+ "step": 116
797
+ },
798
+ {
799
+ "epoch": 0.33,
800
+ "learning_rate": 0.0002940846461288748,
801
+ "loss": 0.8164,
802
+ "step": 117
803
+ },
804
+ {
805
+ "epoch": 0.33,
806
+ "learning_rate": 0.0002939488064366024,
807
+ "loss": 0.6519,
808
+ "step": 118
809
+ },
810
+ {
811
+ "epoch": 0.33,
812
+ "learning_rate": 0.00029381145678946584,
813
+ "loss": 0.6519,
814
+ "step": 119
815
+ },
816
+ {
817
+ "epoch": 0.34,
818
+ "learning_rate": 0.00029367259862819804,
819
+ "loss": 0.7479,
820
+ "step": 120
821
+ },
822
+ {
823
+ "epoch": 0.34,
824
+ "learning_rate": 0.0002935322334093553,
825
+ "loss": 0.5756,
826
+ "step": 121
827
+ },
828
+ {
829
+ "epoch": 0.34,
830
+ "learning_rate": 0.0002933903626053024,
831
+ "loss": 0.3582,
832
+ "step": 122
833
+ },
834
+ {
835
+ "epoch": 0.35,
836
+ "learning_rate": 0.00029324698770419685,
837
+ "loss": 0.5282,
838
+ "step": 123
839
+ },
840
+ {
841
+ "epoch": 0.35,
842
+ "learning_rate": 0.0002931021102099737,
843
+ "loss": 0.5435,
844
+ "step": 124
845
+ },
846
+ {
847
+ "epoch": 0.35,
848
+ "learning_rate": 0.0002929557316423291,
849
+ "loss": 0.7049,
850
+ "step": 125
851
+ },
852
+ {
853
+ "epoch": 0.35,
854
+ "eval_loss": 0.539155125617981,
855
+ "eval_runtime": 26.1963,
856
+ "eval_samples_per_second": 7.635,
857
+ "eval_steps_per_second": 1.909,
858
+ "step": 125
859
+ },
860
+ {
861
+ "dharma_eval_accuracy": 0.25951081518986185,
862
+ "dharma_eval_accuracy_ARC-Challenge": 0.42592592592592593,
863
+ "dharma_eval_accuracy_ARC-Easy": 0.46296296296296297,
864
+ "dharma_eval_accuracy_BoolQ": 0.1111111111111111,
865
+ "dharma_eval_accuracy_MMLU": 0.4074074074074074,
866
+ "dharma_eval_accuracy_agieval": 0.2542372881355932,
867
+ "dharma_eval_accuracy_bigbench": 0.22950819672131148,
868
+ "dharma_eval_accuracy_openbookqa": 0.14814814814814814,
869
+ "dharma_eval_accuracy_truthful_qa": 0.2962962962962963,
870
+ "dharma_eval_accuracy_winogrande": 0.0,
871
+ "dharma_loss": 2.083115490436554,
872
+ "epoch": 0.35,
873
+ "step": 125
874
+ },
875
+ {
876
+ "epoch": 0.35,
877
+ "learning_rate": 0.00029280785353670514,
878
+ "loss": 0.5792,
879
+ "step": 126
880
+ },
881
+ {
882
+ "epoch": 0.36,
883
+ "learning_rate": 0.00029265847744427303,
884
+ "loss": 0.8134,
885
+ "step": 127
886
+ },
887
+ {
888
+ "epoch": 0.36,
889
+ "learning_rate": 0.00029250760493191734,
890
+ "loss": 0.6023,
891
+ "step": 128
892
+ },
893
+ {
894
+ "epoch": 0.36,
895
+ "learning_rate": 0.0002923552375822194,
896
+ "loss": 0.6515,
897
+ "step": 129
898
+ },
899
+ {
900
+ "epoch": 0.36,
901
+ "learning_rate": 0.00029220137699344055,
902
+ "loss": 0.7513,
903
+ "step": 130
904
+ },
905
+ {
906
+ "epoch": 0.37,
907
+ "learning_rate": 0.00029204602477950556,
908
+ "loss": 0.6878,
909
+ "step": 131
910
+ },
911
+ {
912
+ "epoch": 0.37,
913
+ "learning_rate": 0.00029188918256998564,
914
+ "loss": 0.4415,
915
+ "step": 132
916
+ },
917
+ {
918
+ "epoch": 0.37,
919
+ "learning_rate": 0.0002917308520100814,
920
+ "loss": 0.5476,
921
+ "step": 133
922
+ },
923
+ {
924
+ "epoch": 0.38,
925
+ "learning_rate": 0.00029157103476060544,
926
+ "loss": 0.5394,
927
+ "step": 134
928
+ },
929
+ {
930
+ "epoch": 0.38,
931
+ "learning_rate": 0.0002914097324979651,
932
+ "loss": 0.4909,
933
+ "step": 135
934
+ },
935
+ {
936
+ "epoch": 0.38,
937
+ "learning_rate": 0.0002912469469141448,
938
+ "loss": 0.4633,
939
+ "step": 136
940
+ },
941
+ {
942
+ "epoch": 0.38,
943
+ "learning_rate": 0.0002910826797166882,
944
+ "loss": 0.6073,
945
+ "step": 137
946
+ },
947
+ {
948
+ "epoch": 0.39,
949
+ "learning_rate": 0.00029091693262868067,
950
+ "loss": 0.7209,
951
+ "step": 138
952
+ },
953
+ {
954
+ "epoch": 0.39,
955
+ "learning_rate": 0.0002907497073887305,
956
+ "loss": 0.6085,
957
+ "step": 139
958
+ },
959
+ {
960
+ "epoch": 0.39,
961
+ "learning_rate": 0.0002905810057509515,
962
+ "loss": 0.7864,
963
+ "step": 140
964
+ },
965
+ {
966
+ "epoch": 0.4,
967
+ "learning_rate": 0.00029041082948494395,
968
+ "loss": 0.5827,
969
+ "step": 141
970
+ },
971
+ {
972
+ "epoch": 0.4,
973
+ "learning_rate": 0.00029023918037577635,
974
+ "loss": 0.5981,
975
+ "step": 142
976
+ },
977
+ {
978
+ "epoch": 0.4,
979
+ "learning_rate": 0.00029006606022396667,
980
+ "loss": 0.5293,
981
+ "step": 143
982
+ },
983
+ {
984
+ "epoch": 0.4,
985
+ "learning_rate": 0.00028989147084546333,
986
+ "loss": 0.7339,
987
+ "step": 144
988
+ },
989
+ {
990
+ "epoch": 0.41,
991
+ "learning_rate": 0.00028971541407162634,
992
+ "loss": 0.6057,
993
+ "step": 145
994
+ },
995
+ {
996
+ "epoch": 0.41,
997
+ "learning_rate": 0.0002895378917492079,
998
+ "loss": 0.5364,
999
+ "step": 146
1000
+ },
1001
+ {
1002
+ "epoch": 0.41,
1003
+ "learning_rate": 0.0002893589057403332,
1004
+ "loss": 0.5716,
1005
+ "step": 147
1006
+ },
1007
+ {
1008
+ "epoch": 0.42,
1009
+ "learning_rate": 0.0002891784579224808,
1010
+ "loss": 1.1362,
1011
+ "step": 148
1012
+ },
1013
+ {
1014
+ "epoch": 0.42,
1015
+ "learning_rate": 0.00028899655018846295,
1016
+ "loss": 0.5743,
1017
+ "step": 149
1018
+ },
1019
+ {
1020
+ "epoch": 0.42,
1021
+ "learning_rate": 0.0002888131844464056,
1022
+ "loss": 0.5567,
1023
+ "step": 150
1024
+ },
1025
+ {
1026
+ "epoch": 0.42,
1027
+ "eval_loss": 0.5322738885879517,
1028
+ "eval_runtime": 26.2089,
1029
+ "eval_samples_per_second": 7.631,
1030
+ "eval_steps_per_second": 1.908,
1031
+ "step": 150
1032
+ },
1033
+ {
1034
+ "dharma_eval_accuracy": 0.38818510400122574,
1035
+ "dharma_eval_accuracy_ARC-Challenge": 0.4074074074074074,
1036
+ "dharma_eval_accuracy_ARC-Easy": 0.46296296296296297,
1037
+ "dharma_eval_accuracy_BoolQ": 0.7037037037037037,
1038
+ "dharma_eval_accuracy_MMLU": 0.37037037037037035,
1039
+ "dharma_eval_accuracy_agieval": 0.1864406779661017,
1040
+ "dharma_eval_accuracy_bigbench": 0.3442622950819672,
1041
+ "dharma_eval_accuracy_openbookqa": 0.2037037037037037,
1042
+ "dharma_eval_accuracy_truthful_qa": 0.4074074074074074,
1043
+ "dharma_eval_accuracy_winogrande": 0.4074074074074074,
1044
+ "dharma_loss": 2.3776819763183594,
1045
+ "epoch": 0.42,
1046
+ "step": 150
1047
+ },
1048
+ {
1049
+ "epoch": 0.42,
1050
+ "learning_rate": 0.0002886283626197287,
1051
+ "loss": 0.6541,
1052
+ "step": 151
1053
+ },
1054
+ {
1055
+ "epoch": 0.43,
1056
+ "learning_rate": 0.00028844208664712575,
1057
+ "loss": 0.4995,
1058
+ "step": 152
1059
+ },
1060
+ {
1061
+ "epoch": 0.43,
1062
+ "learning_rate": 0.0002882543584825435,
1063
+ "loss": 0.4582,
1064
+ "step": 153
1065
+ },
1066
+ {
1067
+ "epoch": 0.43,
1068
+ "learning_rate": 0.00028806518009516155,
1069
+ "loss": 0.7142,
1070
+ "step": 154
1071
+ },
1072
+ {
1073
+ "epoch": 0.44,
1074
+ "learning_rate": 0.0002878745534693718,
1075
+ "loss": 0.5899,
1076
+ "step": 155
1077
+ },
1078
+ {
1079
+ "epoch": 0.44,
1080
+ "learning_rate": 0.00028768248060475726,
1081
+ "loss": 0.4097,
1082
+ "step": 156
1083
+ },
1084
+ {
1085
+ "epoch": 0.44,
1086
+ "learning_rate": 0.0002874889635160714,
1087
+ "loss": 0.5437,
1088
+ "step": 157
1089
+ },
1090
+ {
1091
+ "epoch": 0.44,
1092
+ "learning_rate": 0.0002872940042332169,
1093
+ "loss": 0.5694,
1094
+ "step": 158
1095
+ },
1096
+ {
1097
+ "epoch": 0.45,
1098
+ "learning_rate": 0.0002870976048012244,
1099
+ "loss": 0.5558,
1100
+ "step": 159
1101
+ },
1102
+ {
1103
+ "epoch": 0.45,
1104
+ "learning_rate": 0.000286899767280231,
1105
+ "loss": 0.6174,
1106
+ "step": 160
1107
+ },
1108
+ {
1109
+ "epoch": 0.45,
1110
+ "learning_rate": 0.0002867004937454587,
1111
+ "loss": 0.4809,
1112
+ "step": 161
1113
+ },
1114
+ {
1115
+ "epoch": 0.45,
1116
+ "learning_rate": 0.00028649978628719254,
1117
+ "loss": 0.7249,
1118
+ "step": 162
1119
+ },
1120
+ {
1121
+ "epoch": 0.46,
1122
+ "learning_rate": 0.00028629764701075885,
1123
+ "loss": 0.5729,
1124
+ "step": 163
1125
+ },
1126
+ {
1127
+ "epoch": 0.46,
1128
+ "learning_rate": 0.0002860940780365029,
1129
+ "loss": 0.3942,
1130
+ "step": 164
1131
+ },
1132
+ {
1133
+ "epoch": 0.46,
1134
+ "learning_rate": 0.000285889081499767,
1135
+ "loss": 0.62,
1136
+ "step": 165
1137
+ },
1138
+ {
1139
+ "epoch": 0.47,
1140
+ "learning_rate": 0.00028568265955086774,
1141
+ "loss": 0.6196,
1142
+ "step": 166
1143
+ },
1144
+ {
1145
+ "epoch": 0.47,
1146
+ "learning_rate": 0.0002854748143550738,
1147
+ "loss": 0.4698,
1148
+ "step": 167
1149
+ },
1150
+ {
1151
+ "epoch": 0.47,
1152
+ "learning_rate": 0.00028526554809258275,
1153
+ "loss": 0.53,
1154
+ "step": 168
1155
+ },
1156
+ {
1157
+ "epoch": 0.47,
1158
+ "learning_rate": 0.0002850548629584988,
1159
+ "loss": 0.6017,
1160
+ "step": 169
1161
+ },
1162
+ {
1163
+ "epoch": 0.48,
1164
+ "learning_rate": 0.00028484276116280926,
1165
+ "loss": 0.5452,
1166
+ "step": 170
1167
+ },
1168
+ {
1169
+ "epoch": 0.48,
1170
+ "learning_rate": 0.00028462924493036164,
1171
+ "loss": 0.5359,
1172
+ "step": 171
1173
+ },
1174
+ {
1175
+ "epoch": 0.48,
1176
+ "learning_rate": 0.00028441431650084016,
1177
+ "loss": 0.4914,
1178
+ "step": 172
1179
+ },
1180
+ {
1181
+ "epoch": 0.49,
1182
+ "learning_rate": 0.00028419797812874237,
1183
+ "loss": 0.44,
1184
+ "step": 173
1185
+ },
1186
+ {
1187
+ "epoch": 0.49,
1188
+ "learning_rate": 0.00028398023208335533,
1189
+ "loss": 0.4378,
1190
+ "step": 174
1191
+ },
1192
+ {
1193
+ "epoch": 0.49,
1194
+ "learning_rate": 0.00028376108064873215,
1195
+ "loss": 0.5328,
1196
+ "step": 175
1197
+ },
1198
+ {
1199
+ "epoch": 0.49,
1200
+ "eval_loss": 0.5265315771102905,
1201
+ "eval_runtime": 26.1903,
1202
+ "eval_samples_per_second": 7.636,
1203
+ "eval_steps_per_second": 1.909,
1204
+ "step": 175
1205
+ },
1206
+ {
1207
+ "dharma_eval_accuracy": 0.35997424982019466,
1208
+ "dharma_eval_accuracy_ARC-Challenge": 0.4444444444444444,
1209
+ "dharma_eval_accuracy_ARC-Easy": 0.5185185185185185,
1210
+ "dharma_eval_accuracy_BoolQ": 0.5925925925925926,
1211
+ "dharma_eval_accuracy_MMLU": 0.3888888888888889,
1212
+ "dharma_eval_accuracy_agieval": 0.22033898305084745,
1213
+ "dharma_eval_accuracy_bigbench": 0.2786885245901639,
1214
+ "dharma_eval_accuracy_openbookqa": 0.1111111111111111,
1215
+ "dharma_eval_accuracy_truthful_qa": 0.3333333333333333,
1216
+ "dharma_eval_accuracy_winogrande": 0.35185185185185186,
1217
+ "dharma_loss": 2.2151358013153075,
1218
+ "epoch": 0.49,
1219
+ "step": 175
1220
+ },
1221
+ {
1222
+ "epoch": 0.49,
1223
+ "learning_rate": 0.00028354052612366757,
1224
+ "loss": 0.7281,
1225
+ "step": 176
1226
+ },
1227
+ {
1228
+ "epoch": 0.5,
1229
+ "learning_rate": 0.0002833185708216743,
1230
+ "loss": 0.6805,
1231
+ "step": 177
1232
+ },
1233
+ {
1234
+ "epoch": 0.5,
1235
+ "learning_rate": 0.0002830952170709583,
1236
+ "loss": 0.5105,
1237
+ "step": 178
1238
+ },
1239
+ {
1240
+ "epoch": 0.5,
1241
+ "learning_rate": 0.00028287046721439484,
1242
+ "loss": 0.565,
1243
+ "step": 179
1244
+ },
1245
+ {
1246
+ "epoch": 0.51,
1247
+ "learning_rate": 0.00028264432360950353,
1248
+ "loss": 0.6562,
1249
+ "step": 180
1250
+ },
1251
+ {
1252
+ "epoch": 0.51,
1253
+ "learning_rate": 0.0002824167886284237,
1254
+ "loss": 0.4868,
1255
+ "step": 181
1256
+ },
1257
+ {
1258
+ "epoch": 0.51,
1259
+ "learning_rate": 0.0002821878646578898,
1260
+ "loss": 0.722,
1261
+ "step": 182
1262
+ },
1263
+ {
1264
+ "epoch": 0.51,
1265
+ "learning_rate": 0.0002819575540992058,
1266
+ "loss": 0.6236,
1267
+ "step": 183
1268
+ },
1269
+ {
1270
+ "epoch": 0.52,
1271
+ "learning_rate": 0.00028172585936822055,
1272
+ "loss": 0.6328,
1273
+ "step": 184
1274
+ },
1275
+ {
1276
+ "epoch": 0.52,
1277
+ "learning_rate": 0.00028149278289530213,
1278
+ "loss": 0.5526,
1279
+ "step": 185
1280
+ },
1281
+ {
1282
+ "epoch": 0.52,
1283
+ "learning_rate": 0.00028125832712531246,
1284
+ "loss": 0.6678,
1285
+ "step": 186
1286
+ },
1287
+ {
1288
+ "epoch": 0.52,
1289
+ "learning_rate": 0.0002810224945175816,
1290
+ "loss": 0.6193,
1291
+ "step": 187
1292
+ },
1293
+ {
1294
+ "epoch": 0.53,
1295
+ "learning_rate": 0.00028078528754588205,
1296
+ "loss": 0.5589,
1297
+ "step": 188
1298
+ },
1299
+ {
1300
+ "epoch": 0.53,
1301
+ "learning_rate": 0.00028054670869840265,
1302
+ "loss": 0.5679,
1303
+ "step": 189
1304
+ },
1305
+ {
1306
+ "epoch": 0.53,
1307
+ "learning_rate": 0.00028030676047772265,
1308
+ "loss": 0.4394,
1309
+ "step": 190
1310
+ },
1311
+ {
1312
+ "epoch": 0.54,
1313
+ "learning_rate": 0.00028006544540078534,
1314
+ "loss": 0.7148,
1315
+ "step": 191
1316
+ },
1317
+ {
1318
+ "epoch": 0.54,
1319
+ "learning_rate": 0.0002798227659988717,
1320
+ "loss": 0.7248,
1321
+ "step": 192
1322
+ },
1323
+ {
1324
+ "epoch": 0.54,
1325
+ "learning_rate": 0.00027957872481757374,
1326
+ "loss": 0.6304,
1327
+ "step": 193
1328
+ },
1329
+ {
1330
+ "epoch": 0.54,
1331
+ "learning_rate": 0.00027933332441676807,
1332
+ "loss": 0.5778,
1333
+ "step": 194
1334
+ },
1335
+ {
1336
+ "epoch": 0.55,
1337
+ "learning_rate": 0.0002790865673705888,
1338
+ "loss": 0.4473,
1339
+ "step": 195
1340
+ },
1341
+ {
1342
+ "epoch": 0.55,
1343
+ "learning_rate": 0.00027883845626740044,
1344
+ "loss": 0.5202,
1345
+ "step": 196
1346
+ },
1347
+ {
1348
+ "epoch": 0.55,
1349
+ "learning_rate": 0.0002785889937097712,
1350
+ "loss": 0.4966,
1351
+ "step": 197
1352
+ },
1353
+ {
1354
+ "epoch": 0.56,
1355
+ "learning_rate": 0.00027833818231444516,
1356
+ "loss": 0.6239,
1357
+ "step": 198
1358
+ },
1359
+ {
1360
+ "epoch": 0.56,
1361
+ "learning_rate": 0.00027808602471231524,
1362
+ "loss": 0.5887,
1363
+ "step": 199
1364
+ },
1365
+ {
1366
+ "epoch": 0.56,
1367
+ "learning_rate": 0.0002778325235483954,
1368
+ "loss": 0.5363,
1369
+ "step": 200
1370
+ },
1371
+ {
1372
+ "epoch": 0.56,
1373
+ "eval_loss": 0.5244021415710449,
1374
+ "eval_runtime": 26.2022,
1375
+ "eval_samples_per_second": 7.633,
1376
+ "eval_steps_per_second": 1.908,
1377
+ "step": 200
1378
+ },
1379
+ {
1380
+ "dharma_eval_accuracy": 0.4286398713863134,
1381
+ "dharma_eval_accuracy_ARC-Challenge": 0.5185185185185185,
1382
+ "dharma_eval_accuracy_ARC-Easy": 0.5370370370370371,
1383
+ "dharma_eval_accuracy_BoolQ": 0.7777777777777778,
1384
+ "dharma_eval_accuracy_MMLU": 0.3888888888888889,
1385
+ "dharma_eval_accuracy_agieval": 0.2542372881355932,
1386
+ "dharma_eval_accuracy_bigbench": 0.3442622950819672,
1387
+ "dharma_eval_accuracy_openbookqa": 0.1111111111111111,
1388
+ "dharma_eval_accuracy_truthful_qa": 0.4444444444444444,
1389
+ "dharma_eval_accuracy_winogrande": 0.48148148148148145,
1390
+ "dharma_loss": 2.3920813593864443,
1391
+ "epoch": 0.56,
1392
+ "step": 200
1393
+ },
1394
+ {
1395
+ "epoch": 0.56,
1396
+ "learning_rate": 0.00027757768148179275,
1397
+ "loss": 0.4511,
1398
+ "step": 201
1399
+ },
1400
+ {
1401
+ "epoch": 0.57,
1402
+ "learning_rate": 0.00027732150118568017,
1403
+ "loss": 0.5601,
1404
+ "step": 202
1405
+ },
1406
+ {
1407
+ "epoch": 0.57,
1408
+ "learning_rate": 0.00027706398534726757,
1409
+ "loss": 0.4554,
1410
+ "step": 203
1411
+ },
1412
+ {
1413
+ "epoch": 0.57,
1414
+ "learning_rate": 0.00027680513666777435,
1415
+ "loss": 0.472,
1416
+ "step": 204
1417
+ },
1418
+ {
1419
+ "epoch": 0.58,
1420
+ "learning_rate": 0.00027654495786240066,
1421
+ "loss": 0.464,
1422
+ "step": 205
1423
+ },
1424
+ {
1425
+ "epoch": 0.58,
1426
+ "learning_rate": 0.00027628345166029906,
1427
+ "loss": 0.526,
1428
+ "step": 206
1429
+ },
1430
+ {
1431
+ "epoch": 0.58,
1432
+ "learning_rate": 0.00027602062080454577,
1433
+ "loss": 0.6065,
1434
+ "step": 207
1435
+ },
1436
+ {
1437
+ "epoch": 0.58,
1438
+ "learning_rate": 0.0002757564680521122,
1439
+ "loss": 0.5417,
1440
+ "step": 208
1441
+ },
1442
+ {
1443
+ "epoch": 0.59,
1444
+ "learning_rate": 0.0002754909961738357,
1445
+ "loss": 0.6115,
1446
+ "step": 209
1447
+ },
1448
+ {
1449
+ "epoch": 0.59,
1450
+ "learning_rate": 0.00027522420795439065,
1451
+ "loss": 0.7071,
1452
+ "step": 210
1453
+ },
1454
+ {
1455
+ "epoch": 0.59,
1456
+ "learning_rate": 0.0002749561061922592,
1457
+ "loss": 0.5509,
1458
+ "step": 211
1459
+ },
1460
+ {
1461
+ "epoch": 0.6,
1462
+ "learning_rate": 0.00027468669369970207,
1463
+ "loss": 0.5462,
1464
+ "step": 212
1465
+ },
1466
+ {
1467
+ "epoch": 0.6,
1468
+ "learning_rate": 0.0002744159733027287,
1469
+ "loss": 0.6335,
1470
+ "step": 213
1471
+ },
1472
+ {
1473
+ "epoch": 0.6,
1474
+ "learning_rate": 0.0002741439478410681,
1475
+ "loss": 0.6087,
1476
+ "step": 214
1477
+ },
1478
+ {
1479
+ "epoch": 0.6,
1480
+ "learning_rate": 0.0002738706201681384,
1481
+ "loss": 0.6883,
1482
+ "step": 215
1483
+ },
1484
+ {
1485
+ "epoch": 0.61,
1486
+ "learning_rate": 0.00027359599315101787,
1487
+ "loss": 0.5886,
1488
+ "step": 216
1489
+ },
1490
+ {
1491
+ "epoch": 0.61,
1492
+ "learning_rate": 0.00027332006967041366,
1493
+ "loss": 0.6278,
1494
+ "step": 217
1495
+ },
1496
+ {
1497
+ "epoch": 0.61,
1498
+ "learning_rate": 0.0002730428526206327,
1499
+ "loss": 0.4986,
1500
+ "step": 218
1501
+ },
1502
+ {
1503
+ "epoch": 0.61,
1504
+ "learning_rate": 0.0002727643449095507,
1505
+ "loss": 0.7309,
1506
+ "step": 219
1507
+ },
1508
+ {
1509
+ "epoch": 0.62,
1510
+ "learning_rate": 0.0002724845494585816,
1511
+ "loss": 0.6104,
1512
+ "step": 220
1513
+ },
1514
+ {
1515
+ "epoch": 0.62,
1516
+ "learning_rate": 0.0002722034692026474,
1517
+ "loss": 0.745,
1518
+ "step": 221
1519
+ },
1520
+ {
1521
+ "epoch": 0.62,
1522
+ "learning_rate": 0.00027192110709014697,
1523
+ "loss": 0.549,
1524
+ "step": 222
1525
+ },
1526
+ {
1527
+ "epoch": 0.63,
1528
+ "learning_rate": 0.0002716374660829252,
1529
+ "loss": 0.563,
1530
+ "step": 223
1531
+ },
1532
+ {
1533
+ "epoch": 0.63,
1534
+ "learning_rate": 0.0002713525491562421,
1535
+ "loss": 0.6249,
1536
+ "step": 224
1537
+ },
1538
+ {
1539
+ "epoch": 0.63,
1540
+ "learning_rate": 0.00027106635929874135,
1541
+ "loss": 0.6111,
1542
+ "step": 225
1543
+ },
1544
+ {
1545
+ "epoch": 0.63,
1546
+ "eval_loss": 0.5195571184158325,
1547
+ "eval_runtime": 26.3108,
1548
+ "eval_samples_per_second": 7.601,
1549
+ "eval_steps_per_second": 1.9,
1550
+ "step": 225
1551
+ },
1552
+ {
1553
+ "dharma_eval_accuracy": 0.4018291546462952,
1554
+ "dharma_eval_accuracy_ARC-Challenge": 0.5740740740740741,
1555
+ "dharma_eval_accuracy_ARC-Easy": 0.6296296296296297,
1556
+ "dharma_eval_accuracy_BoolQ": 0.7407407407407407,
1557
+ "dharma_eval_accuracy_MMLU": 0.37037037037037035,
1558
+ "dharma_eval_accuracy_agieval": 0.23728813559322035,
1559
+ "dharma_eval_accuracy_bigbench": 0.36065573770491804,
1560
+ "dharma_eval_accuracy_openbookqa": 0.09259259259259259,
1561
+ "dharma_eval_accuracy_truthful_qa": 0.2777777777777778,
1562
+ "dharma_eval_accuracy_winogrande": 0.3333333333333333,
1563
+ "dharma_loss": 2.2894758434295652,
1564
+ "epoch": 0.63,
1565
+ "step": 225
1566
+ },
1567
+ {
1568
+ "epoch": 0.63,
1569
+ "learning_rate": 0.0002707788995124192,
1570
+ "loss": 0.6386,
1571
+ "step": 226
1572
+ },
1573
+ {
1574
+ "epoch": 0.64,
1575
+ "learning_rate": 0.00027049017281259276,
1576
+ "loss": 0.5923,
1577
+ "step": 227
1578
+ },
1579
+ {
1580
+ "epoch": 0.64,
1581
+ "learning_rate": 0.0002702001822278685,
1582
+ "loss": 0.411,
1583
+ "step": 228
1584
+ },
1585
+ {
1586
+ "epoch": 0.64,
1587
+ "learning_rate": 0.00026990893080011035,
1588
+ "loss": 0.5807,
1589
+ "step": 229
1590
+ },
1591
+ {
1592
+ "epoch": 0.65,
1593
+ "learning_rate": 0.0002696164215844081,
1594
+ "loss": 0.5582,
1595
+ "step": 230
1596
+ },
1597
+ {
1598
+ "epoch": 0.65,
1599
+ "learning_rate": 0.00026932265764904493,
1600
+ "loss": 0.5324,
1601
+ "step": 231
1602
+ },
1603
+ {
1604
+ "epoch": 0.65,
1605
+ "learning_rate": 0.0002690276420754655,
1606
+ "loss": 0.4203,
1607
+ "step": 232
1608
+ },
1609
+ {
1610
+ "epoch": 0.65,
1611
+ "learning_rate": 0.0002687313779582436,
1612
+ "loss": 0.5673,
1613
+ "step": 233
1614
+ },
1615
+ {
1616
+ "epoch": 0.66,
1617
+ "learning_rate": 0.0002684338684050497,
1618
+ "loss": 0.6918,
1619
+ "step": 234
1620
+ },
1621
+ {
1622
+ "epoch": 0.66,
1623
+ "learning_rate": 0.00026813511653661816,
1624
+ "loss": 0.5747,
1625
+ "step": 235
1626
+ },
1627
+ {
1628
+ "epoch": 0.66,
1629
+ "learning_rate": 0.0002678351254867147,
1630
+ "loss": 0.589,
1631
+ "step": 236
1632
+ },
1633
+ {
1634
+ "epoch": 0.67,
1635
+ "learning_rate": 0.0002675338984021035,
1636
+ "loss": 0.2718,
1637
+ "step": 237
1638
+ },
1639
+ {
1640
+ "epoch": 0.67,
1641
+ "learning_rate": 0.00026723143844251416,
1642
+ "loss": 0.6372,
1643
+ "step": 238
1644
+ },
1645
+ {
1646
+ "epoch": 0.67,
1647
+ "learning_rate": 0.0002669277487806085,
1648
+ "loss": 0.4913,
1649
+ "step": 239
1650
+ },
1651
+ {
1652
+ "epoch": 0.67,
1653
+ "learning_rate": 0.0002666228326019474,
1654
+ "loss": 0.5037,
1655
+ "step": 240
1656
+ },
1657
+ {
1658
+ "epoch": 0.68,
1659
+ "learning_rate": 0.0002663166931049572,
1660
+ "loss": 0.5732,
1661
+ "step": 241
1662
+ },
1663
+ {
1664
+ "epoch": 0.68,
1665
+ "learning_rate": 0.00026600933350089654,
1666
+ "loss": 0.5557,
1667
+ "step": 242
1668
+ },
1669
+ {
1670
+ "epoch": 0.68,
1671
+ "learning_rate": 0.00026570075701382213,
1672
+ "loss": 0.7603,
1673
+ "step": 243
1674
+ },
1675
+ {
1676
+ "epoch": 0.68,
1677
+ "learning_rate": 0.0002653909668805553,
1678
+ "loss": 0.6386,
1679
+ "step": 244
1680
+ },
1681
+ {
1682
+ "epoch": 0.69,
1683
+ "learning_rate": 0.00026507996635064787,
1684
+ "loss": 0.5421,
1685
+ "step": 245
1686
+ },
1687
+ {
1688
+ "epoch": 0.69,
1689
+ "learning_rate": 0.0002647677586863484,
1690
+ "loss": 0.6066,
1691
+ "step": 246
1692
+ },
1693
+ {
1694
+ "epoch": 0.69,
1695
+ "learning_rate": 0.0002644543471625675,
1696
+ "loss": 0.6434,
1697
+ "step": 247
1698
+ },
1699
+ {
1700
+ "epoch": 0.7,
1701
+ "learning_rate": 0.00026413973506684364,
1702
+ "loss": 0.5883,
1703
+ "step": 248
1704
+ },
1705
+ {
1706
+ "epoch": 0.7,
1707
+ "learning_rate": 0.000263823925699309,
1708
+ "loss": 0.5793,
1709
+ "step": 249
1710
+ },
1711
+ {
1712
+ "epoch": 0.7,
1713
+ "learning_rate": 0.00026350692237265427,
1714
+ "loss": 0.5622,
1715
+ "step": 250
1716
+ },
1717
+ {
1718
+ "epoch": 0.7,
1719
+ "eval_loss": 0.5192391276359558,
1720
+ "eval_runtime": 26.26,
1721
+ "eval_samples_per_second": 7.616,
1722
+ "eval_steps_per_second": 1.904,
1723
+ "step": 250
1724
+ },
1725
+ {
1726
+ "dharma_eval_accuracy": 0.37803139189326707,
1727
+ "dharma_eval_accuracy_ARC-Challenge": 0.5,
1728
+ "dharma_eval_accuracy_ARC-Easy": 0.5185185185185185,
1729
+ "dharma_eval_accuracy_BoolQ": 0.7592592592592593,
1730
+ "dharma_eval_accuracy_MMLU": 0.42592592592592593,
1731
+ "dharma_eval_accuracy_agieval": 0.288135593220339,
1732
+ "dharma_eval_accuracy_bigbench": 0.26229508196721313,
1733
+ "dharma_eval_accuracy_openbookqa": 0.09259259259259259,
1734
+ "dharma_eval_accuracy_truthful_qa": 0.3333333333333333,
1735
+ "dharma_eval_accuracy_winogrande": 0.2222222222222222,
1736
+ "dharma_loss": 2.7471679282188415,
1737
+ "epoch": 0.7,
1738
+ "step": 250
1739
+ },
1740
+ {
1741
+ "epoch": 0.7,
1742
+ "learning_rate": 0.00026318872841209445,
1743
+ "loss": 0.361,
1744
+ "step": 251
1745
+ },
1746
+ {
1747
+ "epoch": 0.71,
1748
+ "learning_rate": 0.0002628693471553335,
1749
+ "loss": 0.5178,
1750
+ "step": 252
1751
+ },
1752
+ {
1753
+ "epoch": 0.71,
1754
+ "learning_rate": 0.00026254878195252983,
1755
+ "loss": 0.4308,
1756
+ "step": 253
1757
+ },
1758
+ {
1759
+ "epoch": 0.71,
1760
+ "learning_rate": 0.00026222703616626055,
1761
+ "loss": 0.4397,
1762
+ "step": 254
1763
+ },
1764
+ {
1765
+ "epoch": 0.72,
1766
+ "learning_rate": 0.00026190411317148686,
1767
+ "loss": 0.5298,
1768
+ "step": 255
1769
+ },
1770
+ {
1771
+ "epoch": 0.72,
1772
+ "learning_rate": 0.00026158001635551815,
1773
+ "loss": 0.6231,
1774
+ "step": 256
1775
+ },
1776
+ {
1777
+ "epoch": 0.72,
1778
+ "learning_rate": 0.00026125474911797657,
1779
+ "loss": 0.6699,
1780
+ "step": 257
1781
+ },
1782
+ {
1783
+ "epoch": 0.72,
1784
+ "learning_rate": 0.0002609283148707616,
1785
+ "loss": 0.5838,
1786
+ "step": 258
1787
+ },
1788
+ {
1789
+ "epoch": 0.73,
1790
+ "learning_rate": 0.000260600717038014,
1791
+ "loss": 0.8865,
1792
+ "step": 259
1793
+ },
1794
+ {
1795
+ "epoch": 0.73,
1796
+ "learning_rate": 0.00026027195905608006,
1797
+ "loss": 0.4901,
1798
+ "step": 260
1799
+ },
1800
+ {
1801
+ "epoch": 0.73,
1802
+ "learning_rate": 0.00025994204437347537,
1803
+ "loss": 0.5792,
1804
+ "step": 261
1805
+ },
1806
+ {
1807
+ "epoch": 0.74,
1808
+ "learning_rate": 0.00025961097645084885,
1809
+ "loss": 0.6365,
1810
+ "step": 262
1811
+ },
1812
+ {
1813
+ "epoch": 0.74,
1814
+ "learning_rate": 0.0002592787587609465,
1815
+ "loss": 0.4429,
1816
+ "step": 263
1817
+ },
1818
+ {
1819
+ "epoch": 0.74,
1820
+ "learning_rate": 0.00025894539478857445,
1821
+ "loss": 0.6505,
1822
+ "step": 264
1823
+ },
1824
+ {
1825
+ "epoch": 0.74,
1826
+ "learning_rate": 0.00025861088803056325,
1827
+ "loss": 0.4626,
1828
+ "step": 265
1829
+ },
1830
+ {
1831
+ "epoch": 0.75,
1832
+ "learning_rate": 0.0002582752419957303,
1833
+ "loss": 0.4521,
1834
+ "step": 266
1835
+ },
1836
+ {
1837
+ "epoch": 0.75,
1838
+ "learning_rate": 0.0002579384602048438,
1839
+ "loss": 0.5703,
1840
+ "step": 267
1841
+ },
1842
+ {
1843
+ "epoch": 0.75,
1844
+ "learning_rate": 0.00025760054619058535,
1845
+ "loss": 0.6705,
1846
+ "step": 268
1847
+ },
1848
+ {
1849
+ "epoch": 0.76,
1850
+ "learning_rate": 0.00025726150349751305,
1851
+ "loss": 0.5355,
1852
+ "step": 269
1853
+ },
1854
+ {
1855
+ "epoch": 0.76,
1856
+ "learning_rate": 0.0002569213356820244,
1857
+ "loss": 0.6653,
1858
+ "step": 270
1859
+ },
1860
+ {
1861
+ "epoch": 0.76,
1862
+ "learning_rate": 0.0002565800463123187,
1863
+ "loss": 0.6273,
1864
+ "step": 271
1865
+ },
1866
+ {
1867
+ "epoch": 0.76,
1868
+ "learning_rate": 0.0002562376389683599,
1869
+ "loss": 0.5154,
1870
+ "step": 272
1871
+ },
1872
+ {
1873
+ "epoch": 0.77,
1874
+ "learning_rate": 0.00025589411724183923,
1875
+ "loss": 0.5249,
1876
+ "step": 273
1877
+ },
1878
+ {
1879
+ "epoch": 0.77,
1880
+ "learning_rate": 0.0002555494847361369,
1881
+ "loss": 0.5534,
1882
+ "step": 274
1883
+ },
1884
+ {
1885
+ "epoch": 0.77,
1886
+ "learning_rate": 0.0002552037450662849,
1887
+ "loss": 0.5484,
1888
+ "step": 275
1889
+ },
1890
+ {
1891
+ "epoch": 0.77,
1892
+ "eval_loss": 0.5245066285133362,
1893
+ "eval_runtime": 26.2644,
1894
+ "eval_samples_per_second": 7.615,
1895
+ "eval_steps_per_second": 1.904,
1896
+ "step": 275
1897
+ },
1898
+ {
1899
+ "dharma_eval_accuracy": 0.41176847249521764,
1900
+ "dharma_eval_accuracy_ARC-Challenge": 0.5740740740740741,
1901
+ "dharma_eval_accuracy_ARC-Easy": 0.5370370370370371,
1902
+ "dharma_eval_accuracy_BoolQ": 0.7037037037037037,
1903
+ "dharma_eval_accuracy_MMLU": 0.42592592592592593,
1904
+ "dharma_eval_accuracy_agieval": 0.2711864406779661,
1905
+ "dharma_eval_accuracy_bigbench": 0.36065573770491804,
1906
+ "dharma_eval_accuracy_openbookqa": 0.09259259259259259,
1907
+ "dharma_eval_accuracy_truthful_qa": 0.35185185185185186,
1908
+ "dharma_eval_accuracy_winogrande": 0.3888888888888889,
1909
+ "dharma_loss": 2.512154397010803,
1910
+ "epoch": 0.77,
1911
+ "step": 275
1912
+ },
1913
+ {
1914
+ "epoch": 0.77,
1915
+ "learning_rate": 0.0002548569018589286,
1916
+ "loss": 0.5051,
1917
+ "step": 276
1918
+ },
1919
+ {
1920
+ "epoch": 0.78,
1921
+ "learning_rate": 0.0002545089587522893,
1922
+ "loss": 0.6745,
1923
+ "step": 277
1924
+ },
1925
+ {
1926
+ "epoch": 0.78,
1927
+ "learning_rate": 0.0002541599193961254,
1928
+ "loss": 0.5438,
1929
+ "step": 278
1930
+ },
1931
+ {
1932
+ "epoch": 0.78,
1933
+ "learning_rate": 0.0002538097874516947,
1934
+ "loss": 0.5899,
1935
+ "step": 279
1936
+ },
1937
+ {
1938
+ "epoch": 0.79,
1939
+ "learning_rate": 0.00025345856659171563,
1940
+ "loss": 0.4745,
1941
+ "step": 280
1942
+ },
1943
+ {
1944
+ "epoch": 0.79,
1945
+ "learning_rate": 0.0002531062605003287,
1946
+ "loss": 0.4939,
1947
+ "step": 281
1948
+ },
1949
+ {
1950
+ "epoch": 0.79,
1951
+ "learning_rate": 0.00025275287287305814,
1952
+ "loss": 0.487,
1953
+ "step": 282
1954
+ },
1955
+ {
1956
+ "epoch": 0.79,
1957
+ "learning_rate": 0.00025239840741677304,
1958
+ "loss": 0.4682,
1959
+ "step": 283
1960
+ },
1961
+ {
1962
+ "epoch": 0.8,
1963
+ "learning_rate": 0.00025204286784964817,
1964
+ "loss": 0.8083,
1965
+ "step": 284
1966
+ },
1967
+ {
1968
+ "epoch": 0.8,
1969
+ "learning_rate": 0.00025168625790112547,
1970
+ "loss": 0.4625,
1971
+ "step": 285
1972
+ },
1973
+ {
1974
+ "epoch": 0.8,
1975
+ "learning_rate": 0.00025132858131187444,
1976
+ "loss": 0.5517,
1977
+ "step": 286
1978
+ },
1979
+ {
1980
+ "epoch": 0.81,
1981
+ "learning_rate": 0.00025096984183375334,
1982
+ "loss": 0.5384,
1983
+ "step": 287
1984
+ },
1985
+ {
1986
+ "epoch": 0.81,
1987
+ "learning_rate": 0.0002506100432297695,
1988
+ "loss": 0.5904,
1989
+ "step": 288
1990
+ },
1991
+ {
1992
+ "epoch": 0.81,
1993
+ "learning_rate": 0.00025024918927404004,
1994
+ "loss": 0.5956,
1995
+ "step": 289
1996
+ },
1997
+ {
1998
+ "epoch": 0.81,
1999
+ "learning_rate": 0.00024988728375175214,
2000
+ "loss": 0.5965,
2001
+ "step": 290
2002
+ },
2003
+ {
2004
+ "epoch": 0.82,
2005
+ "learning_rate": 0.00024952433045912354,
2006
+ "loss": 0.4771,
2007
+ "step": 291
2008
+ },
2009
+ {
2010
+ "epoch": 0.82,
2011
+ "learning_rate": 0.00024916033320336264,
2012
+ "loss": 0.6467,
2013
+ "step": 292
2014
+ },
2015
+ {
2016
+ "epoch": 0.82,
2017
+ "learning_rate": 0.00024879529580262817,
2018
+ "loss": 0.544,
2019
+ "step": 293
2020
+ },
2021
+ {
2022
+ "epoch": 0.83,
2023
+ "learning_rate": 0.00024842922208598994,
2024
+ "loss": 0.6526,
2025
+ "step": 294
2026
+ },
2027
+ {
2028
+ "epoch": 0.83,
2029
+ "learning_rate": 0.0002480621158933879,
2030
+ "loss": 0.5648,
2031
+ "step": 295
2032
+ },
2033
+ {
2034
+ "epoch": 0.83,
2035
+ "learning_rate": 0.00024769398107559224,
2036
+ "loss": 0.5017,
2037
+ "step": 296
2038
+ },
2039
+ {
2040
+ "epoch": 0.83,
2041
+ "learning_rate": 0.0002473248214941632,
2042
+ "loss": 0.5809,
2043
+ "step": 297
2044
+ },
2045
+ {
2046
+ "epoch": 0.84,
2047
+ "learning_rate": 0.00024695464102141,
2048
+ "loss": 0.4759,
2049
+ "step": 298
2050
+ },
2051
+ {
2052
+ "epoch": 0.84,
2053
+ "learning_rate": 0.0002465834435403506,
2054
+ "loss": 0.4493,
2055
+ "step": 299
2056
+ },
2057
+ {
2058
+ "epoch": 0.84,
2059
+ "learning_rate": 0.00024621123294467096,
2060
+ "loss": 0.5887,
2061
+ "step": 300
2062
+ },
2063
+ {
2064
+ "epoch": 0.84,
2065
+ "eval_loss": 0.5127784609794617,
2066
+ "eval_runtime": 26.3008,
2067
+ "eval_samples_per_second": 7.604,
2068
+ "eval_steps_per_second": 1.901,
2069
+ "step": 300
2070
+ },
2071
+ {
2072
+ "dharma_eval_accuracy": 0.4206398210751272,
2073
+ "dharma_eval_accuracy_ARC-Challenge": 0.48148148148148145,
2074
+ "dharma_eval_accuracy_ARC-Easy": 0.5185185185185185,
2075
+ "dharma_eval_accuracy_BoolQ": 0.7037037037037037,
2076
+ "dharma_eval_accuracy_MMLU": 0.37037037037037035,
2077
+ "dharma_eval_accuracy_agieval": 0.2711864406779661,
2078
+ "dharma_eval_accuracy_bigbench": 0.45901639344262296,
2079
+ "dharma_eval_accuracy_openbookqa": 0.09259259259259259,
2080
+ "dharma_eval_accuracy_truthful_qa": 0.4074074074074074,
2081
+ "dharma_eval_accuracy_winogrande": 0.48148148148148145,
2082
+ "dharma_loss": 2.5652906007766725,
2083
+ "epoch": 0.84,
2084
+ "step": 300
2085
+ },
2086
+ {
2087
+ "epoch": 0.84,
2088
+ "learning_rate": 0.00024583801313868416,
2089
+ "loss": 0.419,
2090
+ "step": 301
2091
+ },
2092
+ {
2093
+ "epoch": 0.85,
2094
+ "learning_rate": 0.0002454637880372892,
2095
+ "loss": 0.4721,
2096
+ "step": 302
2097
+ },
2098
+ {
2099
+ "epoch": 0.85,
2100
+ "learning_rate": 0.0002450885615659305,
2101
+ "loss": 0.4429,
2102
+ "step": 303
2103
+ },
2104
+ {
2105
+ "epoch": 0.85,
2106
+ "learning_rate": 0.00024471233766055605,
2107
+ "loss": 0.6384,
2108
+ "step": 304
2109
+ },
2110
+ {
2111
+ "epoch": 0.86,
2112
+ "learning_rate": 0.00024433512026757664,
2113
+ "loss": 0.6378,
2114
+ "step": 305
2115
+ },
2116
+ {
2117
+ "epoch": 0.86,
2118
+ "learning_rate": 0.00024395691334382427,
2119
+ "loss": 0.5599,
2120
+ "step": 306
2121
+ },
2122
+ {
2123
+ "epoch": 0.86,
2124
+ "learning_rate": 0.00024357772085651055,
2125
+ "loss": 0.4861,
2126
+ "step": 307
2127
+ },
2128
+ {
2129
+ "epoch": 0.86,
2130
+ "learning_rate": 0.00024319754678318528,
2131
+ "loss": 0.4794,
2132
+ "step": 308
2133
+ },
2134
+ {
2135
+ "epoch": 0.87,
2136
+ "learning_rate": 0.00024281639511169456,
2137
+ "loss": 0.394,
2138
+ "step": 309
2139
+ },
2140
+ {
2141
+ "epoch": 0.87,
2142
+ "learning_rate": 0.0002424342698401391,
2143
+ "loss": 0.703,
2144
+ "step": 310
2145
+ },
2146
+ {
2147
+ "epoch": 0.87,
2148
+ "learning_rate": 0.00024205117497683209,
2149
+ "loss": 0.5961,
2150
+ "step": 311
2151
+ },
2152
+ {
2153
+ "epoch": 0.88,
2154
+ "learning_rate": 0.0002416671145402575,
2155
+ "loss": 0.5959,
2156
+ "step": 312
2157
+ },
2158
+ {
2159
+ "epoch": 0.88,
2160
+ "learning_rate": 0.0002412820925590275,
2161
+ "loss": 0.7628,
2162
+ "step": 313
2163
+ },
2164
+ {
2165
+ "epoch": 0.88,
2166
+ "learning_rate": 0.00024089611307184048,
2167
+ "loss": 0.4302,
2168
+ "step": 314
2169
+ },
2170
+ {
2171
+ "epoch": 0.88,
2172
+ "learning_rate": 0.0002405091801274387,
2173
+ "loss": 0.4601,
2174
+ "step": 315
2175
+ },
2176
+ {
2177
+ "epoch": 0.89,
2178
+ "learning_rate": 0.00024012129778456552,
2179
+ "loss": 0.5376,
2180
+ "step": 316
2181
+ },
2182
+ {
2183
+ "epoch": 0.89,
2184
+ "learning_rate": 0.00023973247011192325,
2185
+ "loss": 0.6526,
2186
+ "step": 317
2187
+ },
2188
+ {
2189
+ "epoch": 0.89,
2190
+ "learning_rate": 0.0002393427011881302,
2191
+ "loss": 0.7244,
2192
+ "step": 318
2193
+ },
2194
+ {
2195
+ "epoch": 0.9,
2196
+ "learning_rate": 0.0002389519951016779,
2197
+ "loss": 0.7522,
2198
+ "step": 319
2199
+ },
2200
+ {
2201
+ "epoch": 0.9,
2202
+ "learning_rate": 0.00023856035595088839,
2203
+ "loss": 0.901,
2204
+ "step": 320
2205
+ },
2206
+ {
2207
+ "epoch": 0.9,
2208
+ "learning_rate": 0.00023816778784387094,
2209
+ "loss": 0.6776,
2210
+ "step": 321
2211
+ },
2212
+ {
2213
+ "epoch": 0.9,
2214
+ "learning_rate": 0.00023777429489847934,
2215
+ "loss": 0.6197,
2216
+ "step": 322
2217
+ },
2218
+ {
2219
+ "epoch": 0.91,
2220
+ "learning_rate": 0.0002373798812422683,
2221
+ "loss": 0.3997,
2222
+ "step": 323
2223
+ },
2224
+ {
2225
+ "epoch": 0.91,
2226
+ "learning_rate": 0.0002369845510124505,
2227
+ "loss": 0.3073,
2228
+ "step": 324
2229
+ },
2230
+ {
2231
+ "epoch": 0.91,
2232
+ "learning_rate": 0.0002365883083558529,
2233
+ "loss": 0.5186,
2234
+ "step": 325
2235
+ },
2236
+ {
2237
+ "epoch": 0.91,
2238
+ "eval_loss": 0.5124163627624512,
2239
+ "eval_runtime": 26.2196,
2240
+ "eval_samples_per_second": 7.628,
2241
+ "eval_steps_per_second": 1.907,
2242
+ "step": 325
2243
+ },
2244
+ {
2245
+ "dharma_eval_accuracy": 0.3996080301226793,
2246
+ "dharma_eval_accuracy_ARC-Challenge": 0.5,
2247
+ "dharma_eval_accuracy_ARC-Easy": 0.48148148148148145,
2248
+ "dharma_eval_accuracy_BoolQ": 0.7037037037037037,
2249
+ "dharma_eval_accuracy_MMLU": 0.42592592592592593,
2250
+ "dharma_eval_accuracy_agieval": 0.3220338983050847,
2251
+ "dharma_eval_accuracy_bigbench": 0.3114754098360656,
2252
+ "dharma_eval_accuracy_openbookqa": 0.09259259259259259,
2253
+ "dharma_eval_accuracy_truthful_qa": 0.42592592592592593,
2254
+ "dharma_eval_accuracy_winogrande": 0.3333333333333333,
2255
+ "dharma_loss": 2.396534740447998,
2256
+ "epoch": 0.91,
2257
+ "step": 325
2258
+ },
2259
+ {
2260
+ "epoch": 0.92,
2261
+ "learning_rate": 0.00023619115742887356,
2262
+ "loss": 0.3524,
2263
+ "step": 326
2264
+ },
2265
+ {
2266
+ "epoch": 0.92,
2267
+ "learning_rate": 0.00023579310239743773,
2268
+ "loss": 0.5638,
2269
+ "step": 327
2270
+ },
2271
+ {
2272
+ "epoch": 0.92,
2273
+ "learning_rate": 0.00023539414743695437,
2274
+ "loss": 0.4728,
2275
+ "step": 328
2276
+ },
2277
+ {
2278
+ "epoch": 0.92,
2279
+ "learning_rate": 0.00023499429673227222,
2280
+ "loss": 0.6048,
2281
+ "step": 329
2282
+ },
2283
+ {
2284
+ "epoch": 0.93,
2285
+ "learning_rate": 0.00023459355447763596,
2286
+ "loss": 0.5342,
2287
+ "step": 330
2288
+ },
2289
+ {
2290
+ "epoch": 0.93,
2291
+ "learning_rate": 0.00023419192487664217,
2292
+ "loss": 0.5119,
2293
+ "step": 331
2294
+ },
2295
+ {
2296
+ "epoch": 0.93,
2297
+ "learning_rate": 0.0002337894121421954,
2298
+ "loss": 0.4893,
2299
+ "step": 332
2300
+ },
2301
+ {
2302
+ "epoch": 0.93,
2303
+ "learning_rate": 0.0002333860204964637,
2304
+ "loss": 0.5625,
2305
+ "step": 333
2306
+ },
2307
+ {
2308
+ "epoch": 0.94,
2309
+ "learning_rate": 0.00023298175417083459,
2310
+ "loss": 0.6029,
2311
+ "step": 334
2312
+ },
2313
+ {
2314
+ "epoch": 0.94,
2315
+ "learning_rate": 0.0002325766174058705,
2316
+ "loss": 0.5609,
2317
+ "step": 335
2318
+ },
2319
+ {
2320
+ "epoch": 0.94,
2321
+ "learning_rate": 0.0002321706144512644,
2322
+ "loss": 0.4422,
2323
+ "step": 336
2324
+ },
2325
+ {
2326
+ "epoch": 0.95,
2327
+ "learning_rate": 0.0002317637495657952,
2328
+ "loss": 0.4773,
2329
+ "step": 337
2330
+ },
2331
+ {
2332
+ "epoch": 0.95,
2333
+ "learning_rate": 0.000231356027017283,
2334
+ "loss": 0.5279,
2335
+ "step": 338
2336
+ },
2337
+ {
2338
+ "epoch": 0.95,
2339
+ "learning_rate": 0.00023094745108254435,
2340
+ "loss": 0.5441,
2341
+ "step": 339
2342
+ },
2343
+ {
2344
+ "epoch": 0.95,
2345
+ "learning_rate": 0.00023053802604734757,
2346
+ "loss": 0.6003,
2347
+ "step": 340
2348
+ },
2349
+ {
2350
+ "epoch": 0.96,
2351
+ "learning_rate": 0.00023012775620636746,
2352
+ "loss": 0.5532,
2353
+ "step": 341
2354
+ },
2355
+ {
2356
+ "epoch": 0.96,
2357
+ "learning_rate": 0.00022971664586314054,
2358
+ "loss": 0.6534,
2359
+ "step": 342
2360
+ },
2361
+ {
2362
+ "epoch": 0.96,
2363
+ "learning_rate": 0.00022930469933001975,
2364
+ "loss": 0.5123,
2365
+ "step": 343
2366
+ },
2367
+ {
2368
+ "epoch": 0.97,
2369
+ "learning_rate": 0.00022889192092812938,
2370
+ "loss": 0.4821,
2371
+ "step": 344
2372
+ },
2373
+ {
2374
+ "epoch": 0.97,
2375
+ "learning_rate": 0.0002284783149873195,
2376
+ "loss": 0.6106,
2377
+ "step": 345
2378
+ },
2379
+ {
2380
+ "epoch": 0.97,
2381
+ "learning_rate": 0.00022806388584612066,
2382
+ "loss": 0.6158,
2383
+ "step": 346
2384
+ },
2385
+ {
2386
+ "epoch": 0.97,
2387
+ "learning_rate": 0.00022764863785169854,
2388
+ "loss": 0.595,
2389
+ "step": 347
2390
+ },
2391
+ {
2392
+ "epoch": 0.98,
2393
+ "learning_rate": 0.000227232575359808,
2394
+ "loss": 0.5091,
2395
+ "step": 348
2396
+ },
2397
+ {
2398
+ "epoch": 0.98,
2399
+ "learning_rate": 0.0002268157027347478,
2400
+ "loss": 0.3905,
2401
+ "step": 349
2402
+ },
2403
+ {
2404
+ "epoch": 0.98,
2405
+ "learning_rate": 0.00022639802434931444,
2406
+ "loss": 0.6038,
2407
+ "step": 350
2408
+ },
2409
+ {
2410
+ "epoch": 0.98,
2411
+ "eval_loss": 0.5118417143821716,
2412
+ "eval_runtime": 26.228,
2413
+ "eval_samples_per_second": 7.625,
2414
+ "eval_steps_per_second": 1.906,
2415
+ "step": 350
2416
+ },
2417
+ {
2418
+ "dharma_eval_accuracy": 0.40288626127285015,
2419
+ "dharma_eval_accuracy_ARC-Challenge": 0.4074074074074074,
2420
+ "dharma_eval_accuracy_ARC-Easy": 0.5555555555555556,
2421
+ "dharma_eval_accuracy_BoolQ": 0.7407407407407407,
2422
+ "dharma_eval_accuracy_MMLU": 0.3148148148148148,
2423
+ "dharma_eval_accuracy_agieval": 0.2033898305084746,
2424
+ "dharma_eval_accuracy_bigbench": 0.3114754098360656,
2425
+ "dharma_eval_accuracy_openbookqa": 0.18518518518518517,
2426
+ "dharma_eval_accuracy_truthful_qa": 0.4444444444444444,
2427
+ "dharma_eval_accuracy_winogrande": 0.46296296296296297,
2428
+ "dharma_loss": 2.5560754494667055,
2429
+ "epoch": 0.98,
2430
+ "step": 350
2431
+ },
2432
+ {
2433
+ "epoch": 0.99,
2434
+ "learning_rate": 0.0002259795445847566,
2435
+ "loss": 0.5352,
2436
+ "step": 351
2437
+ },
2438
+ {
2439
+ "epoch": 0.99,
2440
+ "learning_rate": 0.00022556026783072895,
2441
+ "loss": 0.374,
2442
+ "step": 352
2443
+ },
2444
+ {
2445
+ "epoch": 0.99,
2446
+ "learning_rate": 0.00022514019848524628,
2447
+ "loss": 0.5222,
2448
+ "step": 353
2449
+ },
2450
+ {
2451
+ "epoch": 0.99,
2452
+ "learning_rate": 0.00022471934095463723,
2453
+ "loss": 0.4099,
2454
+ "step": 354
2455
+ },
2456
+ {
2457
+ "epoch": 1.0,
2458
+ "learning_rate": 0.00022429769965349815,
2459
+ "loss": 0.4797,
2460
+ "step": 355
2461
+ },
2462
+ {
2463
+ "epoch": 1.0,
2464
+ "learning_rate": 0.00022387527900464673,
2465
+ "loss": 0.454,
2466
+ "step": 356
2467
+ },
2468
+ {
2469
+ "epoch": 1.0,
2470
+ "learning_rate": 0.00022345208343907576,
2471
+ "loss": 0.6351,
2472
+ "step": 357
2473
+ },
2474
+ {
2475
+ "epoch": 1.0,
2476
+ "learning_rate": 0.00022302811739590642,
2477
+ "loss": 0.4919,
2478
+ "step": 358
2479
+ },
2480
+ {
2481
+ "epoch": 1.01,
2482
+ "learning_rate": 0.00022260338532234194,
2483
+ "loss": 0.4066,
2484
+ "step": 359
2485
+ },
2486
+ {
2487
+ "epoch": 1.01,
2488
+ "learning_rate": 0.00022217789167362073,
2489
+ "loss": 0.5947,
2490
+ "step": 360
2491
+ },
2492
+ {
2493
+ "epoch": 1.01,
2494
+ "learning_rate": 0.00022175164091296987,
2495
+ "loss": 0.6658,
2496
+ "step": 361
2497
+ },
2498
+ {
2499
+ "epoch": 1.02,
2500
+ "learning_rate": 0.00022132463751155812,
2501
+ "loss": 0.6098,
2502
+ "step": 362
2503
+ },
2504
+ {
2505
+ "epoch": 1.02,
2506
+ "learning_rate": 0.00022089688594844915,
2507
+ "loss": 0.5961,
2508
+ "step": 363
2509
+ },
2510
+ {
2511
+ "epoch": 1.02,
2512
+ "learning_rate": 0.00022046839071055435,
2513
+ "loss": 0.3676,
2514
+ "step": 364
2515
+ },
2516
+ {
2517
+ "epoch": 1.02,
2518
+ "learning_rate": 0.00022003915629258605,
2519
+ "loss": 0.485,
2520
+ "step": 365
2521
+ },
2522
+ {
2523
+ "epoch": 1.03,
2524
+ "learning_rate": 0.00021960918719701026,
2525
+ "loss": 0.506,
2526
+ "step": 366
2527
+ },
2528
+ {
2529
+ "epoch": 1.03,
2530
+ "learning_rate": 0.00021917848793399924,
2531
+ "loss": 0.4384,
2532
+ "step": 367
2533
+ },
2534
+ {
2535
+ "epoch": 1.03,
2536
+ "learning_rate": 0.00021874706302138448,
2537
+ "loss": 0.5722,
2538
+ "step": 368
2539
+ },
2540
+ {
2541
+ "epoch": 1.04,
2542
+ "learning_rate": 0.00021831491698460923,
2543
+ "loss": 0.4142,
2544
+ "step": 369
2545
+ },
2546
+ {
2547
+ "epoch": 1.04,
2548
+ "learning_rate": 0.00021788205435668083,
2549
+ "loss": 0.4151,
2550
+ "step": 370
2551
+ },
2552
+ {
2553
+ "epoch": 1.04,
2554
+ "learning_rate": 0.0002174484796781235,
2555
+ "loss": 0.4785,
2556
+ "step": 371
2557
+ },
2558
+ {
2559
+ "epoch": 1.04,
2560
+ "learning_rate": 0.00021701419749693034,
2561
+ "loss": 0.406,
2562
+ "step": 372
2563
+ },
2564
+ {
2565
+ "epoch": 1.05,
2566
+ "learning_rate": 0.00021657921236851603,
2567
+ "loss": 0.3883,
2568
+ "step": 373
2569
+ },
2570
+ {
2571
+ "epoch": 1.05,
2572
+ "learning_rate": 0.00021614352885566872,
2573
+ "loss": 0.508,
2574
+ "step": 374
2575
+ },
2576
+ {
2577
+ "epoch": 1.05,
2578
+ "learning_rate": 0.00021570715152850232,
2579
+ "loss": 0.481,
2580
+ "step": 375
2581
+ },
2582
+ {
2583
+ "epoch": 1.05,
2584
+ "eval_loss": 0.5141562223434448,
2585
+ "eval_runtime": 26.3236,
2586
+ "eval_samples_per_second": 7.598,
2587
+ "eval_steps_per_second": 1.899,
2588
+ "step": 375
2589
+ },
2590
+ {
2591
+ "dharma_eval_accuracy": 0.41410737093179745,
2592
+ "dharma_eval_accuracy_ARC-Challenge": 0.24074074074074073,
2593
+ "dharma_eval_accuracy_ARC-Easy": 0.35185185185185186,
2594
+ "dharma_eval_accuracy_BoolQ": 0.7037037037037037,
2595
+ "dharma_eval_accuracy_MMLU": 0.2222222222222222,
2596
+ "dharma_eval_accuracy_agieval": 0.23728813559322035,
2597
+ "dharma_eval_accuracy_bigbench": 0.5081967213114754,
2598
+ "dharma_eval_accuracy_openbookqa": 0.25925925925925924,
2599
+ "dharma_eval_accuracy_truthful_qa": 0.7037037037037037,
2600
+ "dharma_eval_accuracy_winogrande": 0.5,
2601
+ "dharma_loss": 2.894161937713623,
2602
+ "epoch": 1.05,
2603
+ "step": 375
2604
+ },
2605
+ {
2606
+ "epoch": 1.06,
2607
+ "learning_rate": 0.00021527008496440846,
2608
+ "loss": 0.6044,
2609
+ "step": 376
2610
+ },
2611
+ {
2612
+ "epoch": 1.06,
2613
+ "learning_rate": 0.00021483233374800862,
2614
+ "loss": 0.5436,
2615
+ "step": 377
2616
+ },
2617
+ {
2618
+ "epoch": 1.06,
2619
+ "learning_rate": 0.00021439390247110588,
2620
+ "loss": 0.336,
2621
+ "step": 378
2622
+ },
2623
+ {
2624
+ "epoch": 1.06,
2625
+ "learning_rate": 0.00021395479573263688,
2626
+ "loss": 0.6639,
2627
+ "step": 379
2628
+ },
2629
+ {
2630
+ "epoch": 1.07,
2631
+ "learning_rate": 0.00021351501813862356,
2632
+ "loss": 0.4795,
2633
+ "step": 380
2634
+ },
2635
+ {
2636
+ "epoch": 1.07,
2637
+ "learning_rate": 0.00021307457430212468,
2638
+ "loss": 0.4301,
2639
+ "step": 381
2640
+ },
2641
+ {
2642
+ "epoch": 1.07,
2643
+ "learning_rate": 0.00021263346884318777,
2644
+ "loss": 0.6004,
2645
+ "step": 382
2646
+ },
2647
+ {
2648
+ "epoch": 1.08,
2649
+ "learning_rate": 0.00021219170638880037,
2650
+ "loss": 0.3414,
2651
+ "step": 383
2652
+ },
2653
+ {
2654
+ "epoch": 1.08,
2655
+ "learning_rate": 0.00021174929157284156,
2656
+ "loss": 0.3466,
2657
+ "step": 384
2658
+ },
2659
+ {
2660
+ "epoch": 1.08,
2661
+ "learning_rate": 0.00021130622903603344,
2662
+ "loss": 0.4323,
2663
+ "step": 385
2664
+ },
2665
+ {
2666
+ "epoch": 1.08,
2667
+ "learning_rate": 0.0002108625234258923,
2668
+ "loss": 0.4462,
2669
+ "step": 386
2670
+ },
2671
+ {
2672
+ "epoch": 1.09,
2673
+ "learning_rate": 0.00021041817939668005,
2674
+ "loss": 0.5546,
2675
+ "step": 387
2676
+ },
2677
+ {
2678
+ "epoch": 1.09,
2679
+ "learning_rate": 0.00020997320160935534,
2680
+ "loss": 0.3643,
2681
+ "step": 388
2682
+ },
2683
+ {
2684
+ "epoch": 1.09,
2685
+ "learning_rate": 0.00020952759473152457,
2686
+ "loss": 0.3303,
2687
+ "step": 389
2688
+ },
2689
+ {
2690
+ "epoch": 1.09,
2691
+ "learning_rate": 0.00020908136343739307,
2692
+ "loss": 0.4555,
2693
+ "step": 390
2694
+ },
2695
+ {
2696
+ "epoch": 1.1,
2697
+ "learning_rate": 0.00020863451240771592,
2698
+ "loss": 0.4731,
2699
+ "step": 391
2700
+ },
2701
+ {
2702
+ "epoch": 1.1,
2703
+ "learning_rate": 0.00020818704632974896,
2704
+ "loss": 0.4177,
2705
+ "step": 392
2706
+ },
2707
+ {
2708
+ "epoch": 1.1,
2709
+ "learning_rate": 0.00020773896989719965,
2710
+ "loss": 0.4841,
2711
+ "step": 393
2712
+ },
2713
+ {
2714
+ "epoch": 1.11,
2715
+ "learning_rate": 0.00020729028781017777,
2716
+ "loss": 0.4863,
2717
+ "step": 394
2718
+ },
2719
+ {
2720
+ "epoch": 1.11,
2721
+ "learning_rate": 0.000206841004775146,
2722
+ "loss": 0.3087,
2723
+ "step": 395
2724
+ },
2725
+ {
2726
+ "epoch": 1.11,
2727
+ "learning_rate": 0.00020639112550487084,
2728
+ "loss": 0.5821,
2729
+ "step": 396
2730
+ },
2731
+ {
2732
+ "epoch": 1.11,
2733
+ "learning_rate": 0.000205940654718373,
2734
+ "loss": 0.521,
2735
+ "step": 397
2736
+ },
2737
+ {
2738
+ "epoch": 1.12,
2739
+ "learning_rate": 0.0002054895971408778,
2740
+ "loss": 0.4498,
2741
+ "step": 398
2742
+ },
2743
+ {
2744
+ "epoch": 1.12,
2745
+ "learning_rate": 0.0002050379575037658,
2746
+ "loss": 0.4921,
2747
+ "step": 399
2748
+ },
2749
+ {
2750
+ "epoch": 1.12,
2751
+ "learning_rate": 0.00020458574054452313,
2752
+ "loss": 0.4947,
2753
+ "step": 400
2754
+ },
2755
+ {
2756
+ "epoch": 1.12,
2757
+ "eval_loss": 0.5163459777832031,
2758
+ "eval_runtime": 26.3481,
2759
+ "eval_samples_per_second": 7.591,
2760
+ "eval_steps_per_second": 1.898,
2761
+ "step": 400
2762
+ },
2763
+ {
2764
+ "dharma_eval_accuracy": 0.38530078656965755,
2765
+ "dharma_eval_accuracy_ARC-Challenge": 0.2777777777777778,
2766
+ "dharma_eval_accuracy_ARC-Easy": 0.3888888888888889,
2767
+ "dharma_eval_accuracy_BoolQ": 0.7037037037037037,
2768
+ "dharma_eval_accuracy_MMLU": 0.2222222222222222,
2769
+ "dharma_eval_accuracy_agieval": 0.23728813559322035,
2770
+ "dharma_eval_accuracy_bigbench": 0.5081967213114754,
2771
+ "dharma_eval_accuracy_openbookqa": 0.09259259259259259,
2772
+ "dharma_eval_accuracy_truthful_qa": 0.5370370370370371,
2773
+ "dharma_eval_accuracy_winogrande": 0.5,
2774
+ "dharma_loss": 2.720780613183975,
2775
+ "epoch": 1.12,
2776
+ "step": 400
2777
+ },
2778
+ {
2779
+ "epoch": 1.13,
2780
+ "learning_rate": 0.00020413295100669165,
2781
+ "loss": 0.3095,
2782
+ "step": 401
2783
+ },
2784
+ {
2785
+ "epoch": 1.13,
2786
+ "learning_rate": 0.00020367959363981936,
2787
+ "loss": 0.5584,
2788
+ "step": 402
2789
+ },
2790
+ {
2791
+ "epoch": 1.13,
2792
+ "learning_rate": 0.0002032256731994106,
2793
+ "loss": 0.6251,
2794
+ "step": 403
2795
+ },
2796
+ {
2797
+ "epoch": 1.13,
2798
+ "learning_rate": 0.00020277119444687584,
2799
+ "loss": 0.4737,
2800
+ "step": 404
2801
+ },
2802
+ {
2803
+ "epoch": 1.14,
2804
+ "learning_rate": 0.0002023161621494823,
2805
+ "loss": 0.4528,
2806
+ "step": 405
2807
+ },
2808
+ {
2809
+ "epoch": 1.14,
2810
+ "learning_rate": 0.00020186058108030342,
2811
+ "loss": 0.4723,
2812
+ "step": 406
2813
+ },
2814
+ {
2815
+ "epoch": 1.14,
2816
+ "learning_rate": 0.00020140445601816897,
2817
+ "loss": 0.4127,
2818
+ "step": 407
2819
+ },
2820
+ {
2821
+ "epoch": 1.15,
2822
+ "learning_rate": 0.00020094779174761505,
2823
+ "loss": 0.4784,
2824
+ "step": 408
2825
+ },
2826
+ {
2827
+ "epoch": 1.15,
2828
+ "learning_rate": 0.00020049059305883379,
2829
+ "loss": 0.426,
2830
+ "step": 409
2831
+ },
2832
+ {
2833
+ "epoch": 1.15,
2834
+ "learning_rate": 0.0002000328647476231,
2835
+ "loss": 0.6978,
2836
+ "step": 410
2837
+ },
2838
+ {
2839
+ "epoch": 1.15,
2840
+ "learning_rate": 0.00019957461161533628,
2841
+ "loss": 0.3775,
2842
+ "step": 411
2843
+ },
2844
+ {
2845
+ "epoch": 1.16,
2846
+ "learning_rate": 0.00019911583846883197,
2847
+ "loss": 0.4139,
2848
+ "step": 412
2849
+ },
2850
+ {
2851
+ "epoch": 1.16,
2852
+ "learning_rate": 0.00019865655012042335,
2853
+ "loss": 0.4038,
2854
+ "step": 413
2855
+ },
2856
+ {
2857
+ "epoch": 1.16,
2858
+ "learning_rate": 0.00019819675138782784,
2859
+ "loss": 0.3976,
2860
+ "step": 414
2861
+ },
2862
+ {
2863
+ "epoch": 1.16,
2864
+ "learning_rate": 0.0001977364470941166,
2865
+ "loss": 0.4561,
2866
+ "step": 415
2867
+ },
2868
+ {
2869
+ "epoch": 1.17,
2870
+ "learning_rate": 0.0001972756420676638,
2871
+ "loss": 0.5168,
2872
+ "step": 416
2873
+ },
2874
+ {
2875
+ "epoch": 1.17,
2876
+ "learning_rate": 0.00019681434114209614,
2877
+ "loss": 0.6435,
2878
+ "step": 417
2879
+ },
2880
+ {
2881
+ "epoch": 1.17,
2882
+ "learning_rate": 0.0001963525491562421,
2883
+ "loss": 0.406,
2884
+ "step": 418
2885
+ },
2886
+ {
2887
+ "epoch": 1.18,
2888
+ "learning_rate": 0.00019589027095408105,
2889
+ "loss": 0.4184,
2890
+ "step": 419
2891
+ },
2892
+ {
2893
+ "epoch": 1.18,
2894
+ "learning_rate": 0.0001954275113846926,
2895
+ "loss": 0.4418,
2896
+ "step": 420
2897
+ },
2898
+ {
2899
+ "epoch": 1.18,
2900
+ "learning_rate": 0.00019496427530220565,
2901
+ "loss": 0.5278,
2902
+ "step": 421
2903
+ },
2904
+ {
2905
+ "epoch": 1.18,
2906
+ "learning_rate": 0.0001945005675657475,
2907
+ "loss": 0.5906,
2908
+ "step": 422
2909
+ },
2910
+ {
2911
+ "epoch": 1.19,
2912
+ "learning_rate": 0.0001940363930393929,
2913
+ "loss": 0.3454,
2914
+ "step": 423
2915
+ },
2916
+ {
2917
+ "epoch": 1.19,
2918
+ "learning_rate": 0.00019357175659211298,
2919
+ "loss": 0.6147,
2920
+ "step": 424
2921
+ },
2922
+ {
2923
+ "epoch": 1.19,
2924
+ "learning_rate": 0.00019310666309772424,
2925
+ "loss": 0.3444,
2926
+ "step": 425
2927
+ },
2928
+ {
2929
+ "epoch": 1.19,
2930
+ "eval_loss": 0.5132355690002441,
2931
+ "eval_runtime": 26.3222,
2932
+ "eval_samples_per_second": 7.598,
2933
+ "eval_steps_per_second": 1.9,
2934
+ "step": 425
2935
+ },
2936
+ {
2937
+ "dharma_eval_accuracy": 0.40176169191945177,
2938
+ "dharma_eval_accuracy_ARC-Challenge": 0.3148148148148148,
2939
+ "dharma_eval_accuracy_ARC-Easy": 0.4444444444444444,
2940
+ "dharma_eval_accuracy_BoolQ": 0.7037037037037037,
2941
+ "dharma_eval_accuracy_MMLU": 0.2777777777777778,
2942
+ "dharma_eval_accuracy_agieval": 0.23728813559322035,
2943
+ "dharma_eval_accuracy_bigbench": 0.5081967213114754,
2944
+ "dharma_eval_accuracy_openbookqa": 0.1111111111111111,
2945
+ "dharma_eval_accuracy_truthful_qa": 0.5185185185185185,
2946
+ "dharma_eval_accuracy_winogrande": 0.5,
2947
+ "dharma_loss": 2.5127969179153444,
2948
+ "epoch": 1.19,
2949
+ "step": 425
2950
+ },
2951
+ {
2952
+ "epoch": 1.2,
2953
+ "learning_rate": 0.0001926411174348373,
2954
+ "loss": 0.529,
2955
+ "step": 426
2956
+ },
2957
+ {
2958
+ "epoch": 1.2,
2959
+ "learning_rate": 0.00019217512448680584,
2960
+ "loss": 0.4912,
2961
+ "step": 427
2962
+ },
2963
+ {
2964
+ "epoch": 1.2,
2965
+ "learning_rate": 0.0001917086891416755,
2966
+ "loss": 0.3229,
2967
+ "step": 428
2968
+ },
2969
+ {
2970
+ "epoch": 1.2,
2971
+ "learning_rate": 0.00019124181629213226,
2972
+ "loss": 0.628,
2973
+ "step": 429
2974
+ },
2975
+ {
2976
+ "epoch": 1.21,
2977
+ "learning_rate": 0.0001907745108354514,
2978
+ "loss": 0.7983,
2979
+ "step": 430
2980
+ },
2981
+ {
2982
+ "epoch": 1.21,
2983
+ "learning_rate": 0.0001903067776734461,
2984
+ "loss": 0.4891,
2985
+ "step": 431
2986
+ },
2987
+ {
2988
+ "epoch": 1.21,
2989
+ "learning_rate": 0.00018983862171241577,
2990
+ "loss": 0.5963,
2991
+ "step": 432
2992
+ },
2993
+ {
2994
+ "epoch": 1.22,
2995
+ "learning_rate": 0.000189370047863095,
2996
+ "loss": 0.565,
2997
+ "step": 433
2998
+ },
2999
+ {
3000
+ "epoch": 1.22,
3001
+ "learning_rate": 0.00018890106104060174,
3002
+ "loss": 0.5787,
3003
+ "step": 434
3004
+ },
3005
+ {
3006
+ "epoch": 1.22,
3007
+ "learning_rate": 0.0001884316661643858,
3008
+ "loss": 0.4575,
3009
+ "step": 435
3010
+ },
3011
+ {
3012
+ "epoch": 1.22,
3013
+ "learning_rate": 0.00018796186815817743,
3014
+ "loss": 0.4341,
3015
+ "step": 436
3016
+ },
3017
+ {
3018
+ "epoch": 1.23,
3019
+ "learning_rate": 0.00018749167194993528,
3020
+ "loss": 0.4943,
3021
+ "step": 437
3022
+ },
3023
+ {
3024
+ "epoch": 1.23,
3025
+ "learning_rate": 0.0001870210824717951,
3026
+ "loss": 0.4336,
3027
+ "step": 438
3028
+ },
3029
+ {
3030
+ "epoch": 1.23,
3031
+ "learning_rate": 0.0001865501046600179,
3032
+ "loss": 0.3898,
3033
+ "step": 439
3034
+ },
3035
+ {
3036
+ "epoch": 1.24,
3037
+ "learning_rate": 0.00018607874345493805,
3038
+ "loss": 0.513,
3039
+ "step": 440
3040
+ },
3041
+ {
3042
+ "epoch": 1.24,
3043
+ "learning_rate": 0.0001856070038009115,
3044
+ "loss": 0.547,
3045
+ "step": 441
3046
+ },
3047
+ {
3048
+ "epoch": 1.24,
3049
+ "learning_rate": 0.00018513489064626398,
3050
+ "loss": 0.683,
3051
+ "step": 442
3052
+ },
3053
+ {
3054
+ "epoch": 1.24,
3055
+ "learning_rate": 0.00018466240894323918,
3056
+ "loss": 0.4327,
3057
+ "step": 443
3058
+ },
3059
+ {
3060
+ "epoch": 1.25,
3061
+ "learning_rate": 0.00018418956364794653,
3062
+ "loss": 0.4861,
3063
+ "step": 444
3064
+ },
3065
+ {
3066
+ "epoch": 1.25,
3067
+ "learning_rate": 0.00018371635972030942,
3068
+ "loss": 0.4755,
3069
+ "step": 445
3070
+ },
3071
+ {
3072
+ "epoch": 1.25,
3073
+ "learning_rate": 0.00018324280212401314,
3074
+ "loss": 0.516,
3075
+ "step": 446
3076
+ },
3077
+ {
3078
+ "epoch": 1.25,
3079
+ "learning_rate": 0.00018276889582645277,
3080
+ "loss": 0.4033,
3081
+ "step": 447
3082
+ },
3083
+ {
3084
+ "epoch": 1.26,
3085
+ "learning_rate": 0.00018229464579868121,
3086
+ "loss": 0.3444,
3087
+ "step": 448
3088
+ },
3089
+ {
3090
+ "epoch": 1.26,
3091
+ "learning_rate": 0.00018182005701535678,
3092
+ "loss": 0.4507,
3093
+ "step": 449
3094
+ },
3095
+ {
3096
+ "epoch": 1.26,
3097
+ "learning_rate": 0.00018134513445469127,
3098
+ "loss": 0.4974,
3099
+ "step": 450
3100
+ },
3101
+ {
3102
+ "epoch": 1.26,
3103
+ "eval_loss": 0.5145830512046814,
3104
+ "eval_runtime": 26.3311,
3105
+ "eval_samples_per_second": 7.596,
3106
+ "eval_steps_per_second": 1.899,
3107
+ "step": 450
3108
+ },
3109
+ {
3110
+ "dharma_eval_accuracy": 0.44020972903995964,
3111
+ "dharma_eval_accuracy_ARC-Challenge": 0.3148148148148148,
3112
+ "dharma_eval_accuracy_ARC-Easy": 0.4444444444444444,
3113
+ "dharma_eval_accuracy_BoolQ": 0.7037037037037037,
3114
+ "dharma_eval_accuracy_MMLU": 0.2962962962962963,
3115
+ "dharma_eval_accuracy_agieval": 0.2542372881355932,
3116
+ "dharma_eval_accuracy_bigbench": 0.5409836065573771,
3117
+ "dharma_eval_accuracy_openbookqa": 0.2962962962962963,
3118
+ "dharma_eval_accuracy_truthful_qa": 0.6296296296296297,
3119
+ "dharma_eval_accuracy_winogrande": 0.48148148148148145,
3120
+ "dharma_loss": 2.99742707490921,
3121
+ "epoch": 1.26,
3122
+ "step": 450
3123
+ },
3124
+ {
3125
+ "epoch": 1.27,
3126
+ "learning_rate": 0.00018086988309839753,
3127
+ "loss": 0.5276,
3128
+ "step": 451
3129
+ },
3130
+ {
3131
+ "epoch": 1.27,
3132
+ "learning_rate": 0.00018039430793163753,
3133
+ "loss": 0.603,
3134
+ "step": 452
3135
+ },
3136
+ {
3137
+ "epoch": 1.27,
3138
+ "learning_rate": 0.0001799184139429696,
3139
+ "loss": 0.2147,
3140
+ "step": 453
3141
+ },
3142
+ {
3143
+ "epoch": 1.27,
3144
+ "learning_rate": 0.00017944220612429662,
3145
+ "loss": 0.4451,
3146
+ "step": 454
3147
+ },
3148
+ {
3149
+ "epoch": 1.28,
3150
+ "learning_rate": 0.0001789656894708132,
3151
+ "loss": 0.3871,
3152
+ "step": 455
3153
+ },
3154
+ {
3155
+ "epoch": 1.28,
3156
+ "learning_rate": 0.00017848886898095352,
3157
+ "loss": 0.531,
3158
+ "step": 456
3159
+ },
3160
+ {
3161
+ "epoch": 1.28,
3162
+ "learning_rate": 0.00017801174965633893,
3163
+ "loss": 0.3615,
3164
+ "step": 457
3165
+ },
3166
+ {
3167
+ "epoch": 1.29,
3168
+ "learning_rate": 0.00017753433650172538,
3169
+ "loss": 0.4243,
3170
+ "step": 458
3171
+ },
3172
+ {
3173
+ "epoch": 1.29,
3174
+ "learning_rate": 0.00017705663452495098,
3175
+ "loss": 0.5004,
3176
+ "step": 459
3177
+ },
3178
+ {
3179
+ "epoch": 1.29,
3180
+ "learning_rate": 0.00017657864873688343,
3181
+ "loss": 0.5854,
3182
+ "step": 460
3183
+ },
3184
+ {
3185
+ "epoch": 1.29,
3186
+ "learning_rate": 0.00017610038415136734,
3187
+ "loss": 0.597,
3188
+ "step": 461
3189
+ },
3190
+ {
3191
+ "epoch": 1.3,
3192
+ "learning_rate": 0.00017562184578517203,
3193
+ "loss": 0.5124,
3194
+ "step": 462
3195
+ },
3196
+ {
3197
+ "epoch": 1.3,
3198
+ "learning_rate": 0.00017514303865793848,
3199
+ "loss": 0.371,
3200
+ "step": 463
3201
+ },
3202
+ {
3203
+ "epoch": 1.3,
3204
+ "learning_rate": 0.00017466396779212693,
3205
+ "loss": 0.5462,
3206
+ "step": 464
3207
+ },
3208
+ {
3209
+ "epoch": 1.31,
3210
+ "learning_rate": 0.00017418463821296395,
3211
+ "loss": 0.4221,
3212
+ "step": 465
3213
+ },
3214
+ {
3215
+ "epoch": 1.31,
3216
+ "learning_rate": 0.00017370505494839012,
3217
+ "loss": 0.3771,
3218
+ "step": 466
3219
+ },
3220
+ {
3221
+ "epoch": 1.31,
3222
+ "learning_rate": 0.00017322522302900696,
3223
+ "loss": 0.6663,
3224
+ "step": 467
3225
+ },
3226
+ {
3227
+ "epoch": 1.31,
3228
+ "learning_rate": 0.00017274514748802418,
3229
+ "loss": 0.4786,
3230
+ "step": 468
3231
+ },
3232
+ {
3233
+ "epoch": 1.32,
3234
+ "learning_rate": 0.00017226483336120707,
3235
+ "loss": 0.4326,
3236
+ "step": 469
3237
+ },
3238
+ {
3239
+ "epoch": 1.32,
3240
+ "learning_rate": 0.00017178428568682353,
3241
+ "loss": 0.5531,
3242
+ "step": 470
3243
+ },
3244
+ {
3245
+ "epoch": 1.32,
3246
+ "learning_rate": 0.00017130350950559132,
3247
+ "loss": 0.449,
3248
+ "step": 471
3249
+ },
3250
+ {
3251
+ "epoch": 1.32,
3252
+ "learning_rate": 0.00017082250986062502,
3253
+ "loss": 0.4973,
3254
+ "step": 472
3255
+ },
3256
+ {
3257
+ "epoch": 1.33,
3258
+ "learning_rate": 0.00017034129179738328,
3259
+ "loss": 0.4068,
3260
+ "step": 473
3261
+ },
3262
+ {
3263
+ "epoch": 1.33,
3264
+ "learning_rate": 0.00016985986036361598,
3265
+ "loss": 0.3231,
3266
+ "step": 474
3267
+ },
3268
+ {
3269
+ "epoch": 1.33,
3270
+ "learning_rate": 0.00016937822060931096,
3271
+ "loss": 0.4542,
3272
+ "step": 475
3273
+ },
3274
+ {
3275
+ "epoch": 1.33,
3276
+ "eval_loss": 0.5115104913711548,
3277
+ "eval_runtime": 26.308,
3278
+ "eval_samples_per_second": 7.602,
3279
+ "eval_steps_per_second": 1.901,
3280
+ "step": 475
3281
+ },
3282
+ {
3283
+ "dharma_eval_accuracy": 0.43056827628159167,
3284
+ "dharma_eval_accuracy_ARC-Challenge": 0.46296296296296297,
3285
+ "dharma_eval_accuracy_ARC-Easy": 0.4444444444444444,
3286
+ "dharma_eval_accuracy_BoolQ": 0.7037037037037037,
3287
+ "dharma_eval_accuracy_MMLU": 0.2962962962962963,
3288
+ "dharma_eval_accuracy_agieval": 0.23728813559322035,
3289
+ "dharma_eval_accuracy_bigbench": 0.5081967213114754,
3290
+ "dharma_eval_accuracy_openbookqa": 0.2037037037037037,
3291
+ "dharma_eval_accuracy_truthful_qa": 0.5555555555555556,
3292
+ "dharma_eval_accuracy_winogrande": 0.46296296296296297,
3293
+ "dharma_loss": 3.081619900226593,
3294
+ "epoch": 1.33,
3295
+ "step": 475
3296
+ },
3297
+ {
3298
+ "epoch": 1.34,
3299
+ "learning_rate": 0.00016889637758664137,
3300
+ "loss": 0.3903,
3301
+ "step": 476
3302
+ },
3303
+ {
3304
+ "epoch": 1.34,
3305
+ "learning_rate": 0.0001684143363499125,
3306
+ "loss": 0.4442,
3307
+ "step": 477
3308
+ },
3309
+ {
3310
+ "epoch": 1.34,
3311
+ "learning_rate": 0.00016793210195550882,
3312
+ "loss": 0.484,
3313
+ "step": 478
3314
+ },
3315
+ {
3316
+ "epoch": 1.34,
3317
+ "learning_rate": 0.00016744967946184097,
3318
+ "loss": 0.4949,
3319
+ "step": 479
3320
+ },
3321
+ {
3322
+ "epoch": 1.35,
3323
+ "learning_rate": 0.00016696707392929266,
3324
+ "loss": 0.5704,
3325
+ "step": 480
3326
+ },
3327
+ {
3328
+ "epoch": 1.35,
3329
+ "learning_rate": 0.0001664842904201674,
3330
+ "loss": 0.4988,
3331
+ "step": 481
3332
+ },
3333
+ {
3334
+ "epoch": 1.35,
3335
+ "learning_rate": 0.00016600133399863594,
3336
+ "loss": 0.2734,
3337
+ "step": 482
3338
+ },
3339
+ {
3340
+ "epoch": 1.36,
3341
+ "learning_rate": 0.00016551820973068255,
3342
+ "loss": 0.3876,
3343
+ "step": 483
3344
+ },
3345
+ {
3346
+ "epoch": 1.36,
3347
+ "learning_rate": 0.00016503492268405215,
3348
+ "loss": 0.4982,
3349
+ "step": 484
3350
+ },
3351
+ {
3352
+ "epoch": 1.36,
3353
+ "learning_rate": 0.0001645514779281972,
3354
+ "loss": 0.5982,
3355
+ "step": 485
3356
+ },
3357
+ {
3358
+ "epoch": 1.36,
3359
+ "learning_rate": 0.00016406788053422445,
3360
+ "loss": 0.3784,
3361
+ "step": 486
3362
+ },
3363
+ {
3364
+ "epoch": 1.37,
3365
+ "learning_rate": 0.00016358413557484167,
3366
+ "loss": 0.4585,
3367
+ "step": 487
3368
+ },
3369
+ {
3370
+ "epoch": 1.37,
3371
+ "learning_rate": 0.00016310024812430458,
3372
+ "loss": 0.5089,
3373
+ "step": 488
3374
+ },
3375
+ {
3376
+ "epoch": 1.37,
3377
+ "learning_rate": 0.00016261622325836362,
3378
+ "loss": 0.338,
3379
+ "step": 489
3380
+ },
3381
+ {
3382
+ "epoch": 1.38,
3383
+ "learning_rate": 0.00016213206605421063,
3384
+ "loss": 0.4797,
3385
+ "step": 490
3386
+ },
3387
+ {
3388
+ "epoch": 1.38,
3389
+ "learning_rate": 0.0001616477815904255,
3390
+ "loss": 0.4304,
3391
+ "step": 491
3392
+ },
3393
+ {
3394
+ "epoch": 1.38,
3395
+ "learning_rate": 0.0001611633749469231,
3396
+ "loss": 0.4831,
3397
+ "step": 492
3398
+ },
3399
+ {
3400
+ "epoch": 1.38,
3401
+ "learning_rate": 0.0001606788512048999,
3402
+ "loss": 0.3629,
3403
+ "step": 493
3404
+ },
3405
+ {
3406
+ "epoch": 1.39,
3407
+ "learning_rate": 0.00016019421544678073,
3408
+ "loss": 0.4143,
3409
+ "step": 494
3410
+ },
3411
+ {
3412
+ "epoch": 1.39,
3413
+ "learning_rate": 0.00015970947275616536,
3414
+ "loss": 0.3162,
3415
+ "step": 495
3416
+ },
3417
+ {
3418
+ "epoch": 1.39,
3419
+ "learning_rate": 0.00015922462821777524,
3420
+ "loss": 0.4689,
3421
+ "step": 496
3422
+ },
3423
+ {
3424
+ "epoch": 1.4,
3425
+ "learning_rate": 0.00015873968691740005,
3426
+ "loss": 0.3899,
3427
+ "step": 497
3428
+ },
3429
+ {
3430
+ "epoch": 1.4,
3431
+ "learning_rate": 0.00015825465394184472,
3432
+ "loss": 0.4622,
3433
+ "step": 498
3434
+ },
3435
+ {
3436
+ "epoch": 1.4,
3437
+ "learning_rate": 0.00015776953437887553,
3438
+ "loss": 0.5949,
3439
+ "step": 499
3440
+ },
3441
+ {
3442
+ "epoch": 1.4,
3443
+ "learning_rate": 0.00015728433331716724,
3444
+ "loss": 0.6421,
3445
+ "step": 500
3446
+ },
3447
+ {
3448
+ "epoch": 1.4,
3449
+ "eval_loss": 0.5089285373687744,
3450
+ "eval_runtime": 26.3003,
3451
+ "eval_samples_per_second": 7.604,
3452
+ "eval_steps_per_second": 1.901,
3453
+ "step": 500
3454
+ },
3455
+ {
3456
+ "dharma_eval_accuracy": 0.43468350261904026,
3457
+ "dharma_eval_accuracy_ARC-Challenge": 0.4444444444444444,
3458
+ "dharma_eval_accuracy_ARC-Easy": 0.46296296296296297,
3459
+ "dharma_eval_accuracy_BoolQ": 0.7037037037037037,
3460
+ "dharma_eval_accuracy_MMLU": 0.3333333333333333,
3461
+ "dharma_eval_accuracy_agieval": 0.23728813559322035,
3462
+ "dharma_eval_accuracy_bigbench": 0.5081967213114754,
3463
+ "dharma_eval_accuracy_openbookqa": 0.2037037037037037,
3464
+ "dharma_eval_accuracy_truthful_qa": 0.5555555555555556,
3465
+ "dharma_eval_accuracy_winogrande": 0.46296296296296297,
3466
+ "dharma_loss": 2.98004837846756,
3467
+ "epoch": 1.4,
3468
+ "step": 500
3469
+ },
3470
+ {
3471
+ "epoch": 1.41,
3472
+ "learning_rate": 0.0001567990558462493,
3473
+ "loss": 0.3583,
3474
+ "step": 501
3475
+ },
3476
+ {
3477
+ "epoch": 1.41,
3478
+ "learning_rate": 0.0001563137070564528,
3479
+ "loss": 0.6447,
3480
+ "step": 502
3481
+ },
3482
+ {
3483
+ "epoch": 1.41,
3484
+ "learning_rate": 0.00015582829203885688,
3485
+ "loss": 0.3984,
3486
+ "step": 503
3487
+ },
3488
+ {
3489
+ "epoch": 1.41,
3490
+ "learning_rate": 0.00015534281588523534,
3491
+ "loss": 0.3397,
3492
+ "step": 504
3493
+ },
3494
+ {
3495
+ "epoch": 1.42,
3496
+ "learning_rate": 0.0001548572836880034,
3497
+ "loss": 0.6608,
3498
+ "step": 505
3499
+ },
3500
+ {
3501
+ "epoch": 1.42,
3502
+ "learning_rate": 0.00015437170054016395,
3503
+ "loss": 0.5184,
3504
+ "step": 506
3505
+ },
3506
+ {
3507
+ "epoch": 1.42,
3508
+ "learning_rate": 0.00015388607153525452,
3509
+ "loss": 0.5412,
3510
+ "step": 507
3511
+ },
3512
+ {
3513
+ "epoch": 1.43,
3514
+ "learning_rate": 0.00015340040176729347,
3515
+ "loss": 0.6032,
3516
+ "step": 508
3517
+ },
3518
+ {
3519
+ "epoch": 1.43,
3520
+ "learning_rate": 0.000152914696330727,
3521
+ "loss": 0.5783,
3522
+ "step": 509
3523
+ },
3524
+ {
3525
+ "epoch": 1.43,
3526
+ "learning_rate": 0.00015242896032037522,
3527
+ "loss": 0.5915,
3528
+ "step": 510
3529
+ },
3530
+ {
3531
+ "epoch": 1.43,
3532
+ "learning_rate": 0.00015194319883137905,
3533
+ "loss": 0.6996,
3534
+ "step": 511
3535
+ },
3536
+ {
3537
+ "epoch": 1.44,
3538
+ "learning_rate": 0.0001514574169591466,
3539
+ "loss": 0.3408,
3540
+ "step": 512
3541
+ },
3542
+ {
3543
+ "epoch": 1.44,
3544
+ "learning_rate": 0.0001509716197992999,
3545
+ "loss": 0.4851,
3546
+ "step": 513
3547
+ },
3548
+ {
3549
+ "epoch": 1.44,
3550
+ "learning_rate": 0.0001504858124476213,
3551
+ "loss": 0.3962,
3552
+ "step": 514
3553
+ },
3554
+ {
3555
+ "epoch": 1.45,
3556
+ "learning_rate": 0.00015,
3557
+ "loss": 0.5792,
3558
+ "step": 515
3559
+ },
3560
+ {
3561
+ "epoch": 1.45,
3562
+ "learning_rate": 0.0001495141875523787,
3563
+ "loss": 0.4199,
3564
+ "step": 516
3565
+ },
3566
+ {
3567
+ "epoch": 1.45,
3568
+ "learning_rate": 0.00014902838020070008,
3569
+ "loss": 0.2794,
3570
+ "step": 517
3571
+ },
3572
+ {
3573
+ "epoch": 1.45,
3574
+ "learning_rate": 0.0001485425830408534,
3575
+ "loss": 0.4316,
3576
+ "step": 518
3577
+ },
3578
+ {
3579
+ "epoch": 1.46,
3580
+ "learning_rate": 0.00014805680116862095,
3581
+ "loss": 0.4125,
3582
+ "step": 519
3583
+ },
3584
+ {
3585
+ "epoch": 1.46,
3586
+ "learning_rate": 0.00014757103967962475,
3587
+ "loss": 0.4158,
3588
+ "step": 520
3589
+ },
3590
+ {
3591
+ "epoch": 1.46,
3592
+ "learning_rate": 0.00014708530366927297,
3593
+ "loss": 0.3971,
3594
+ "step": 521
3595
+ },
3596
+ {
3597
+ "epoch": 1.47,
3598
+ "learning_rate": 0.0001465995982327065,
3599
+ "loss": 0.508,
3600
+ "step": 522
3601
+ },
3602
+ {
3603
+ "epoch": 1.47,
3604
+ "learning_rate": 0.0001461139284647455,
3605
+ "loss": 0.4481,
3606
+ "step": 523
3607
+ },
3608
+ {
3609
+ "epoch": 1.47,
3610
+ "learning_rate": 0.00014562829945983607,
3611
+ "loss": 0.5737,
3612
+ "step": 524
3613
+ },
3614
+ {
3615
+ "epoch": 1.47,
3616
+ "learning_rate": 0.00014514271631199663,
3617
+ "loss": 0.4853,
3618
+ "step": 525
3619
+ },
3620
+ {
3621
+ "epoch": 1.47,
3622
+ "eval_loss": 0.508446455001831,
3623
+ "eval_runtime": 26.2523,
3624
+ "eval_samples_per_second": 7.618,
3625
+ "eval_steps_per_second": 1.905,
3626
+ "step": 525
3627
+ },
3628
+ {
3629
+ "dharma_eval_accuracy": 0.4208141950724767,
3630
+ "dharma_eval_accuracy_ARC-Challenge": 0.4444444444444444,
3631
+ "dharma_eval_accuracy_ARC-Easy": 0.46296296296296297,
3632
+ "dharma_eval_accuracy_BoolQ": 0.7037037037037037,
3633
+ "dharma_eval_accuracy_MMLU": 0.35185185185185186,
3634
+ "dharma_eval_accuracy_agieval": 0.2542372881355932,
3635
+ "dharma_eval_accuracy_bigbench": 0.45901639344262296,
3636
+ "dharma_eval_accuracy_openbookqa": 0.16666666666666666,
3637
+ "dharma_eval_accuracy_truthful_qa": 0.5,
3638
+ "dharma_eval_accuracy_winogrande": 0.4444444444444444,
3639
+ "dharma_loss": 2.9239073081016542,
3640
+ "epoch": 1.47,
3641
+ "step": 525
3642
+ },
3643
+ {
3644
+ "epoch": 1.48,
3645
+ "learning_rate": 0.00014465718411476463,
3646
+ "loss": 0.3469,
3647
+ "step": 526
3648
+ },
3649
+ {
3650
+ "epoch": 1.48,
3651
+ "learning_rate": 0.00014417170796114312,
3652
+ "loss": 0.3511,
3653
+ "step": 527
3654
+ },
3655
+ {
3656
+ "epoch": 1.48,
3657
+ "learning_rate": 0.0001436862929435472,
3658
+ "loss": 0.3,
3659
+ "step": 528
3660
+ },
3661
+ {
3662
+ "epoch": 1.48,
3663
+ "learning_rate": 0.00014320094415375072,
3664
+ "loss": 0.4873,
3665
+ "step": 529
3666
+ },
3667
+ {
3668
+ "epoch": 1.49,
3669
+ "learning_rate": 0.0001427156666828328,
3670
+ "loss": 0.3733,
3671
+ "step": 530
3672
+ },
3673
+ {
3674
+ "epoch": 1.49,
3675
+ "learning_rate": 0.00014223046562112442,
3676
+ "loss": 0.5614,
3677
+ "step": 531
3678
+ },
3679
+ {
3680
+ "epoch": 1.49,
3681
+ "learning_rate": 0.00014174534605815525,
3682
+ "loss": 0.5095,
3683
+ "step": 532
3684
+ },
3685
+ {
3686
+ "epoch": 1.5,
3687
+ "learning_rate": 0.00014126031308259992,
3688
+ "loss": 0.5133,
3689
+ "step": 533
3690
+ },
3691
+ {
3692
+ "epoch": 1.5,
3693
+ "learning_rate": 0.0001407753717822248,
3694
+ "loss": 0.6072,
3695
+ "step": 534
3696
+ },
3697
+ {
3698
+ "epoch": 1.5,
3699
+ "learning_rate": 0.00014029052724383464,
3700
+ "loss": 0.4428,
3701
+ "step": 535
3702
+ },
3703
+ {
3704
+ "epoch": 1.5,
3705
+ "learning_rate": 0.00013980578455321922,
3706
+ "loss": 0.4763,
3707
+ "step": 536
3708
+ },
3709
+ {
3710
+ "epoch": 1.51,
3711
+ "learning_rate": 0.00013932114879510005,
3712
+ "loss": 0.4025,
3713
+ "step": 537
3714
+ },
3715
+ {
3716
+ "epoch": 1.51,
3717
+ "learning_rate": 0.0001388366250530769,
3718
+ "loss": 0.5766,
3719
+ "step": 538
3720
+ },
3721
+ {
3722
+ "epoch": 1.51,
3723
+ "learning_rate": 0.0001383522184095745,
3724
+ "loss": 0.6511,
3725
+ "step": 539
3726
+ },
3727
+ {
3728
+ "epoch": 1.52,
3729
+ "learning_rate": 0.00013786793394578937,
3730
+ "loss": 0.481,
3731
+ "step": 540
3732
+ },
3733
+ {
3734
+ "epoch": 1.52,
3735
+ "learning_rate": 0.00013738377674163635,
3736
+ "loss": 0.4244,
3737
+ "step": 541
3738
+ },
3739
+ {
3740
+ "epoch": 1.52,
3741
+ "learning_rate": 0.0001368997518756954,
3742
+ "loss": 0.3272,
3743
+ "step": 542
3744
+ },
3745
+ {
3746
+ "epoch": 1.52,
3747
+ "learning_rate": 0.00013641586442515833,
3748
+ "loss": 0.3597,
3749
+ "step": 543
3750
+ },
3751
+ {
3752
+ "epoch": 1.53,
3753
+ "learning_rate": 0.00013593211946577555,
3754
+ "loss": 0.4322,
3755
+ "step": 544
3756
+ },
3757
+ {
3758
+ "epoch": 1.53,
3759
+ "learning_rate": 0.0001354485220718028,
3760
+ "loss": 0.4576,
3761
+ "step": 545
3762
+ },
3763
+ {
3764
+ "epoch": 1.53,
3765
+ "learning_rate": 0.00013496507731594783,
3766
+ "loss": 0.5028,
3767
+ "step": 546
3768
+ },
3769
+ {
3770
+ "epoch": 1.54,
3771
+ "learning_rate": 0.00013448179026931745,
3772
+ "loss": 0.5469,
3773
+ "step": 547
3774
+ },
3775
+ {
3776
+ "epoch": 1.54,
3777
+ "learning_rate": 0.00013399866600136403,
3778
+ "loss": 0.4966,
3779
+ "step": 548
3780
+ },
3781
+ {
3782
+ "epoch": 1.54,
3783
+ "learning_rate": 0.00013351570957983257,
3784
+ "loss": 0.5205,
3785
+ "step": 549
3786
+ },
3787
+ {
3788
+ "epoch": 1.54,
3789
+ "learning_rate": 0.00013303292607070737,
3790
+ "loss": 0.2469,
3791
+ "step": 550
3792
+ },
3793
+ {
3794
+ "epoch": 1.54,
3795
+ "eval_loss": 0.5087947249412537,
3796
+ "eval_runtime": 26.3254,
3797
+ "eval_samples_per_second": 7.597,
3798
+ "eval_steps_per_second": 1.899,
3799
+ "step": 550
3800
+ },
3801
+ {
3802
+ "dharma_eval_accuracy": 0.4146413555663039,
3803
+ "dharma_eval_accuracy_ARC-Challenge": 0.4444444444444444,
3804
+ "dharma_eval_accuracy_ARC-Easy": 0.5185185185185185,
3805
+ "dharma_eval_accuracy_BoolQ": 0.7037037037037037,
3806
+ "dharma_eval_accuracy_MMLU": 0.3333333333333333,
3807
+ "dharma_eval_accuracy_agieval": 0.2542372881355932,
3808
+ "dharma_eval_accuracy_bigbench": 0.45901639344262296,
3809
+ "dharma_eval_accuracy_openbookqa": 0.12962962962962962,
3810
+ "dharma_eval_accuracy_truthful_qa": 0.42592592592592593,
3811
+ "dharma_eval_accuracy_winogrande": 0.46296296296296297,
3812
+ "dharma_loss": 3.0210403323173525,
3813
+ "epoch": 1.54,
3814
+ "step": 550
3815
+ },
3816
+ {
3817
+ "epoch": 1.55,
3818
+ "learning_rate": 0.000132550320538159,
3819
+ "loss": 0.5321,
3820
+ "step": 551
3821
+ },
3822
+ {
3823
+ "epoch": 1.55,
3824
+ "learning_rate": 0.00013206789804449116,
3825
+ "loss": 0.3957,
3826
+ "step": 552
3827
+ },
3828
+ {
3829
+ "epoch": 1.55,
3830
+ "learning_rate": 0.0001315856636500875,
3831
+ "loss": 0.5538,
3832
+ "step": 553
3833
+ },
3834
+ {
3835
+ "epoch": 1.56,
3836
+ "learning_rate": 0.00013110362241335865,
3837
+ "loss": 0.5354,
3838
+ "step": 554
3839
+ },
3840
+ {
3841
+ "epoch": 1.56,
3842
+ "learning_rate": 0.00013062177939068907,
3843
+ "loss": 0.57,
3844
+ "step": 555
3845
+ },
3846
+ {
3847
+ "epoch": 1.56,
3848
+ "learning_rate": 0.00013014013963638408,
3849
+ "loss": 0.3717,
3850
+ "step": 556
3851
+ },
3852
+ {
3853
+ "epoch": 1.56,
3854
+ "learning_rate": 0.0001296587082026167,
3855
+ "loss": 0.3922,
3856
+ "step": 557
3857
+ },
3858
+ {
3859
+ "epoch": 1.57,
3860
+ "learning_rate": 0.000129177490139375,
3861
+ "loss": 0.3959,
3862
+ "step": 558
3863
+ },
3864
+ {
3865
+ "epoch": 1.57,
3866
+ "learning_rate": 0.0001286964904944087,
3867
+ "loss": 0.4779,
3868
+ "step": 559
3869
+ },
3870
+ {
3871
+ "epoch": 1.57,
3872
+ "learning_rate": 0.00012821571431317647,
3873
+ "loss": 0.4544,
3874
+ "step": 560
3875
+ },
3876
+ {
3877
+ "epoch": 1.57,
3878
+ "learning_rate": 0.0001277351666387929,
3879
+ "loss": 0.3896,
3880
+ "step": 561
3881
+ },
3882
+ {
3883
+ "epoch": 1.58,
3884
+ "learning_rate": 0.0001272548525119758,
3885
+ "loss": 0.3762,
3886
+ "step": 562
3887
+ },
3888
+ {
3889
+ "epoch": 1.58,
3890
+ "learning_rate": 0.000126774776970993,
3891
+ "loss": 0.3926,
3892
+ "step": 563
3893
+ },
3894
+ {
3895
+ "epoch": 1.58,
3896
+ "learning_rate": 0.00012629494505160982,
3897
+ "loss": 0.5463,
3898
+ "step": 564
3899
+ },
3900
+ {
3901
+ "epoch": 1.59,
3902
+ "learning_rate": 0.00012581536178703603,
3903
+ "loss": 0.2925,
3904
+ "step": 565
3905
+ },
3906
+ {
3907
+ "epoch": 1.59,
3908
+ "learning_rate": 0.0001253360322078731,
3909
+ "loss": 0.5089,
3910
+ "step": 566
3911
+ },
3912
+ {
3913
+ "epoch": 1.59,
3914
+ "learning_rate": 0.00012485696134206147,
3915
+ "loss": 0.4544,
3916
+ "step": 567
3917
+ },
3918
+ {
3919
+ "epoch": 1.59,
3920
+ "learning_rate": 0.0001243781542148279,
3921
+ "loss": 0.4922,
3922
+ "step": 568
3923
+ },
3924
+ {
3925
+ "epoch": 1.6,
3926
+ "learning_rate": 0.00012389961584863263,
3927
+ "loss": 0.4084,
3928
+ "step": 569
3929
+ },
3930
+ {
3931
+ "epoch": 1.6,
3932
+ "learning_rate": 0.0001234213512631166,
3933
+ "loss": 0.5996,
3934
+ "step": 570
3935
+ },
3936
+ {
3937
+ "epoch": 1.6,
3938
+ "learning_rate": 0.000122943365475049,
3939
+ "loss": 0.3808,
3940
+ "step": 571
3941
+ },
3942
+ {
3943
+ "epoch": 1.61,
3944
+ "learning_rate": 0.0001224656634982746,
3945
+ "loss": 0.5916,
3946
+ "step": 572
3947
+ },
3948
+ {
3949
+ "epoch": 1.61,
3950
+ "learning_rate": 0.00012198825034366104,
3951
+ "loss": 0.3169,
3952
+ "step": 573
3953
+ },
3954
+ {
3955
+ "epoch": 1.61,
3956
+ "learning_rate": 0.00012151113101904645,
3957
+ "loss": 0.4497,
3958
+ "step": 574
3959
+ },
3960
+ {
3961
+ "epoch": 1.61,
3962
+ "learning_rate": 0.0001210343105291868,
3963
+ "loss": 0.4914,
3964
+ "step": 575
3965
+ },
3966
+ {
3967
+ "epoch": 1.61,
3968
+ "eval_loss": 0.5054534077644348,
3969
+ "eval_runtime": 26.2667,
3970
+ "eval_samples_per_second": 7.614,
3971
+ "eval_steps_per_second": 1.904,
3972
+ "step": 575
3973
+ },
3974
+ {
3975
+ "dharma_eval_accuracy": 0.4225739431506465,
3976
+ "dharma_eval_accuracy_ARC-Challenge": 0.4074074074074074,
3977
+ "dharma_eval_accuracy_ARC-Easy": 0.5,
3978
+ "dharma_eval_accuracy_BoolQ": 0.7037037037037037,
3979
+ "dharma_eval_accuracy_MMLU": 0.3333333333333333,
3980
+ "dharma_eval_accuracy_agieval": 0.23728813559322035,
3981
+ "dharma_eval_accuracy_bigbench": 0.4918032786885246,
3982
+ "dharma_eval_accuracy_openbookqa": 0.16666666666666666,
3983
+ "dharma_eval_accuracy_truthful_qa": 0.5,
3984
+ "dharma_eval_accuracy_winogrande": 0.46296296296296297,
3985
+ "dharma_loss": 2.9983138756752012,
3986
+ "epoch": 1.61,
3987
+ "step": 575
3988
+ },
3989
+ {
3990
+ "epoch": 1.62,
3991
+ "learning_rate": 0.00012055779387570338,
3992
+ "loss": 0.4089,
3993
+ "step": 576
3994
+ },
3995
+ {
3996
+ "epoch": 1.62,
3997
+ "learning_rate": 0.00012008158605703037,
3998
+ "loss": 0.4753,
3999
+ "step": 577
4000
+ },
4001
+ {
4002
+ "epoch": 1.62,
4003
+ "learning_rate": 0.00011960569206836249,
4004
+ "loss": 0.5209,
4005
+ "step": 578
4006
+ },
4007
+ {
4008
+ "epoch": 1.63,
4009
+ "learning_rate": 0.00011913011690160246,
4010
+ "loss": 0.5442,
4011
+ "step": 579
4012
+ },
4013
+ {
4014
+ "epoch": 1.63,
4015
+ "learning_rate": 0.00011865486554530873,
4016
+ "loss": 0.3312,
4017
+ "step": 580
4018
+ },
4019
+ {
4020
+ "epoch": 1.63,
4021
+ "learning_rate": 0.00011817994298464319,
4022
+ "loss": 0.4079,
4023
+ "step": 581
4024
+ },
4025
+ {
4026
+ "epoch": 1.63,
4027
+ "learning_rate": 0.00011770535420131876,
4028
+ "loss": 0.4949,
4029
+ "step": 582
4030
+ },
4031
+ {
4032
+ "epoch": 1.64,
4033
+ "learning_rate": 0.00011723110417354719,
4034
+ "loss": 0.5599,
4035
+ "step": 583
4036
+ },
4037
+ {
4038
+ "epoch": 1.64,
4039
+ "learning_rate": 0.00011675719787598688,
4040
+ "loss": 0.4379,
4041
+ "step": 584
4042
+ },
4043
+ {
4044
+ "epoch": 1.64,
4045
+ "learning_rate": 0.00011628364027969062,
4046
+ "loss": 0.4673,
4047
+ "step": 585
4048
+ },
4049
+ {
4050
+ "epoch": 1.64,
4051
+ "learning_rate": 0.00011581043635205352,
4052
+ "loss": 0.3881,
4053
+ "step": 586
4054
+ },
4055
+ {
4056
+ "epoch": 1.65,
4057
+ "learning_rate": 0.00011533759105676082,
4058
+ "loss": 0.6034,
4059
+ "step": 587
4060
+ },
4061
+ {
4062
+ "epoch": 1.65,
4063
+ "learning_rate": 0.000114865109353736,
4064
+ "loss": 0.4709,
4065
+ "step": 588
4066
+ },
4067
+ {
4068
+ "epoch": 1.65,
4069
+ "learning_rate": 0.00011439299619908854,
4070
+ "loss": 0.5639,
4071
+ "step": 589
4072
+ },
4073
+ {
4074
+ "epoch": 1.66,
4075
+ "learning_rate": 0.00011392125654506198,
4076
+ "loss": 0.4515,
4077
+ "step": 590
4078
+ },
4079
+ {
4080
+ "epoch": 1.66,
4081
+ "learning_rate": 0.00011344989533998204,
4082
+ "loss": 0.4912,
4083
+ "step": 591
4084
+ },
4085
+ {
4086
+ "epoch": 1.66,
4087
+ "learning_rate": 0.00011297891752820484,
4088
+ "loss": 0.4824,
4089
+ "step": 592
4090
+ },
4091
+ {
4092
+ "epoch": 1.66,
4093
+ "learning_rate": 0.0001125083280500647,
4094
+ "loss": 0.6789,
4095
+ "step": 593
4096
+ },
4097
+ {
4098
+ "epoch": 1.67,
4099
+ "learning_rate": 0.00011203813184182255,
4100
+ "loss": 0.3983,
4101
+ "step": 594
4102
+ },
4103
+ {
4104
+ "epoch": 1.67,
4105
+ "learning_rate": 0.00011156833383561413,
4106
+ "loss": 0.5017,
4107
+ "step": 595
4108
+ },
4109
+ {
4110
+ "epoch": 1.67,
4111
+ "learning_rate": 0.00011109893895939823,
4112
+ "loss": 0.3995,
4113
+ "step": 596
4114
+ },
4115
+ {
4116
+ "epoch": 1.68,
4117
+ "learning_rate": 0.00011062995213690494,
4118
+ "loss": 0.6159,
4119
+ "step": 597
4120
+ },
4121
+ {
4122
+ "epoch": 1.68,
4123
+ "learning_rate": 0.00011016137828758422,
4124
+ "loss": 0.5068,
4125
+ "step": 598
4126
+ },
4127
+ {
4128
+ "epoch": 1.68,
4129
+ "learning_rate": 0.0001096932223265539,
4130
+ "loss": 0.6377,
4131
+ "step": 599
4132
+ },
4133
+ {
4134
+ "epoch": 1.68,
4135
+ "learning_rate": 0.00010922548916454855,
4136
+ "loss": 0.527,
4137
+ "step": 600
4138
+ },
4139
+ {
4140
+ "epoch": 1.68,
4141
+ "eval_loss": 0.5055681467056274,
4142
+ "eval_runtime": 26.3047,
4143
+ "eval_samples_per_second": 7.603,
4144
+ "eval_steps_per_second": 1.901,
4145
+ "step": 600
4146
+ },
4147
+ {
4148
+ "dharma_eval_accuracy": 0.41687334273237764,
4149
+ "dharma_eval_accuracy_ARC-Challenge": 0.42592592592592593,
4150
+ "dharma_eval_accuracy_ARC-Easy": 0.5,
4151
+ "dharma_eval_accuracy_BoolQ": 0.7037037037037037,
4152
+ "dharma_eval_accuracy_MMLU": 0.37037037037037035,
4153
+ "dharma_eval_accuracy_agieval": 0.23728813559322035,
4154
+ "dharma_eval_accuracy_bigbench": 0.45901639344262296,
4155
+ "dharma_eval_accuracy_openbookqa": 0.12962962962962962,
4156
+ "dharma_eval_accuracy_truthful_qa": 0.46296296296296297,
4157
+ "dharma_eval_accuracy_winogrande": 0.46296296296296297,
4158
+ "dharma_loss": 2.924492097854614,
4159
+ "epoch": 1.68,
4160
+ "step": 600
4161
+ }
4162
+ ],
4163
+ "max_steps": 1000,
4164
+ "num_train_epochs": 3,
4165
+ "total_flos": 1.9307426648459674e+17,
4166
+ "trial_name": null,
4167
+ "trial_params": null
4168
+ }
checkpoint-600/training_args.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:8d82b859a9cd8d6d889fcf83059a0117f69472b64660a7ed0075051324482a44
3
+ size 5947