Upload folder using huggingface_hub

config.json

@@ -1,5 +1,5 @@
 {
-  "_name_or_path": "/nm/drive0/sadkins/.cache/huggingface/hub/models--TinyLlama--TinyLlama-1.1B-intermediate-step-1431k-3T/snapshots/036fa4651240b9a1487f709833b9e4b96b4c1574",
+  "_name_or_path": "/root/.cache/huggingface/hub/models--TinyLlama--TinyLlama-1.1B-intermediate-step-1431k-3T/snapshots/036fa4651240b9a1487f709833b9e4b96b4c1574",
   "architectures": [
     "LlamaForCausalLM"
   ],
@@ -23,7 +23,12 @@
   "quantization_config": {
     "config_groups": {
       "group_0": {
-        "input_activations": {
+        "input_activations": null,
+        "output_activations": null,
+        "targets": [
+          "Embedding"
+        ],
+        "weights": {
           "block_structure": null,
           "group_size": null,
           "num_bits": 8,
@@ -32,6 +37,18 @@
           "strategy": "tensor",
           "symmetric": true,
           "type": "int"
+        }
+      },
+      "group_1": {
+        "input_activations": {
+          "block_structure": null,
+          "group_size": null,
+          "num_bits": 8,
+          "observer": "minmax",
+          "observer_kwargs": {},
+          "strategy": "tensor",
+          "symmetric": false,
+          "type": "int"
         },
         "output_activations": null,
         "targets": [
@@ -50,7 +67,7 @@
       }
     },
     "format": "fakequant",
-    "global_compression_ratio": 1.4375595368095078,
+    "global_compression_ratio": 1.4416825559554713,
     "ignore": [
       "model.layers.0.mlp.down_proj",
       "lm_head"

model.safetensors

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:836e22041658a79b3024c37b093b0fa02e7f8a3480c306c3491e3353d2e5bf00
-size 4400287762
+oid sha256:3f94cc1d9b92c63daae79a819a45fbe669bbf1afaba13435ab0f0de7aa47130f
+size 4400287663

recipe.yaml

@@ -5,6 +5,17 @@ test_stage:
       config_groups:
         group_0:
           weights: {num_bits: 8, type: int, symmetric: true, strategy: tensor}
-          input_activations: {num_bits: 8, type: int, symmetric: true, strategy: tensor}
+          input_activations: {num_bits: 8, type: int, symmetric: false, strategy: tensor}
           output_activations: null
           targets: [Linear]
+        group_1:
+          weights: {num_bits: 8, type: int, symmetric: true, strategy: tensor}
+          input_activations: null
+          output_activations: null
+          targets: [Embedding]
+    SparseGPTModifier:
+      sparsity: 0.0
+      block_size: 128
+      sequential_update: false
+      quantize: true
+      targets: ['re:model.layers.\d+$']