Spaces:
Runtime error
Runtime error
File size: 36,880 Bytes
4014562 |
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 158 159 160 161 162 163 164 165 166 167 168 169 170 171 172 173 174 175 176 177 178 179 180 181 182 183 184 185 186 187 188 189 190 191 192 193 194 195 196 197 198 199 200 201 202 203 204 205 206 207 208 209 210 211 212 213 214 215 216 217 218 219 220 221 222 223 224 225 226 227 228 229 230 231 232 233 234 235 236 237 238 239 240 241 242 243 244 245 246 247 248 249 250 251 252 253 254 255 256 257 258 259 260 261 262 263 264 265 266 267 268 269 270 271 272 273 274 275 276 277 278 279 280 281 282 283 284 285 286 287 288 289 290 291 292 293 294 295 296 297 298 299 300 301 302 303 304 305 306 307 308 309 310 |
huggingface/tokenizers: The current process just got forked, after parallelism has already been used. Disabling parallelism to avoid deadlocks...
To disable this warning, you can either:
- Avoid using `tokenizers` before the fork if possible
- Explicitly set the environment variable TOKENIZERS_PARALLELISM=(true | false)
huggingface/tokenizers: The current process just got forked, after parallelism has already been used. Disabling parallelism to avoid deadlocks...
To disable this warning, you can either:
- Avoid using `tokenizers` before the fork if possible
- Explicitly set the environment variable TOKENIZERS_PARALLELISM=(true | false)
huggingface/tokenizers: The current process just got forked, after parallelism has already been used. Disabling parallelism to avoid deadlocks...
To disable this warning, you can either:
- Avoid using `tokenizers` before the fork if possible
- Explicitly set the environment variable TOKENIZERS_PARALLELISM=(true | false)
[34m[1mwandb[39m[22m: [33mWARNING[39m Serializing object of type dict that is 589920 bytes
[34m[1mwandb[39m[22m: [33mWARNING[39m Serializing object of type dict that is 589920 bytes
0%| | 0/70340 [00:00<?, ?it/s]
huggingface/tokenizers: The current process just got forked, after parallelism has already been used. Disabling parallelism to avoid deadlocks...
To disable this warning, you can either:
- Avoid using `tokenizers` before the fork if possible
- Explicitly set the environment variable TOKENIZERS_PARALLELISM=(true | false)
huggingface/tokenizers: The current process just got forked, after parallelism has already been used. Disabling parallelism to avoid deadlocks...
To disable this warning, you can either:
- Avoid using `tokenizers` before the fork if possible
- Explicitly set the environment variable TOKENIZERS_PARALLELISM=(true | false)
huggingface/tokenizers: The current process just got forked, after parallelism has already been used. Disabling parallelism to avoid deadlocks...
To disable this warning, you can either:
- Avoid using `tokenizers` before the fork if possible
- Explicitly set the environment variable TOKENIZERS_PARALLELISM=(true | false)
huggingface/tokenizers: The current process just got forked, after parallelism has already been used. Disabling parallelism to avoid deadlocks...
To disable this warning, you can either:
- Avoid using `tokenizers` before the fork if possible
- Explicitly set the environment variable TOKENIZERS_PARALLELISM=(true | false)
huggingface/tokenizers: The current process just got forked, after parallelism has already been used. Disabling parallelism to avoid deadlocks...
To disable this warning, you can either:
- Avoid using `tokenizers` before the fork if possible
- Explicitly set the environment variable TOKENIZERS_PARALLELISM=(true | false)
huggingface/tokenizers: The current process just got forked, after parallelism has already been used. Disabling parallelism to avoid deadlocks...
To disable this warning, you can either:
- Avoid using `tokenizers` before the fork if possible
- Explicitly set the environment variable TOKENIZERS_PARALLELISM=(true | false)
huggingface/tokenizers: The current process just got forked, after parallelism has already been used. Disabling parallelism to avoid deadlocks...
To disable this warning, you can either:
- Avoid using `tokenizers` before the fork if possible
- Explicitly set the environment variable TOKENIZERS_PARALLELISM=(true | false)
huggingface/tokenizers: The current process just got forked, after parallelism has already been used. Disabling parallelism to avoid deadlocks...
To disable this warning, you can either:
- Avoid using `tokenizers` before the fork if possible
- Explicitly set the environment variable TOKENIZERS_PARALLELISM=(true | false)
huggingface/tokenizers: The current process just got forked, after parallelism has already been used. Disabling parallelism to avoid deadlocks...
To disable this warning, you can either:
- Avoid using `tokenizers` before the fork if possible
- Explicitly set the environment variable TOKENIZERS_PARALLELISM=(true | false)
huggingface/tokenizers: The current process just got forked, after parallelism has already been used. Disabling parallelism to avoid deadlocks...
To disable this warning, you can either:
- Avoid using `tokenizers` before the fork if possible
- Explicitly set the environment variable TOKENIZERS_PARALLELISM=(true | false)
huggingface/tokenizers: The current process just got forked, after parallelism has already been used. Disabling parallelism to avoid deadlocks...
To disable this warning, you can either:
- Avoid using `tokenizers` before the fork if possible
- Explicitly set the environment variable TOKENIZERS_PARALLELISM=(true | false)
huggingface/tokenizers: The current process just got forked, after parallelism has already been used. Disabling parallelism to avoid deadlocks...
To disable this warning, you can either:
- Avoid using `tokenizers` before the fork if possible
- Explicitly set the environment variable TOKENIZERS_PARALLELISM=(true | false)
huggingface/tokenizers: The current process just got forked, after parallelism has already been used. Disabling parallelism to avoid deadlocks...
To disable this warning, you can either:
- Avoid using `tokenizers` before the fork if possible
0%| | 0/70340 [00:00<?, ?it/s]/n/fs/nlp-pranjal/miniconda3/lib/python3.9/site-packages/transformers/data/data_collator.py:132: UserWarning: Creating a tensor from a list of numpy.ndarrays is extremely slow. Please consider converting the list to a single numpy.ndarray with numpy.array() before converting to a tensor. (Triggered internally at ../torch/csrc/utils/tensor_new.cpp:210.)
batch[k] = torch.tensor([f[k] for f in features])
/n/fs/nlp-pranjal/miniconda3/lib/python3.9/site-packages/transformers/data/data_collator.py:132: UserWarning: Creating a tensor from a list of numpy.ndarrays is extremely slow. Please consider converting the list to a single numpy.ndarray with numpy.array() before converting to a tensor. (Triggered internally at ../torch/csrc/utils/tensor_new.cpp:210.)
batch[k] = torch.tensor([f[k] for f in features])
/n/fs/nlp-pranjal/miniconda3/lib/python3.9/site-packages/transformers/data/data_collator.py:132: UserWarning: Creating a tensor from a list of numpy.ndarrays is extremely slow. Please consider converting the list to a single numpy.ndarray with numpy.array() before converting to a tensor. (Triggered internally at ../torch/csrc/utils/tensor_new.cpp:210.)
batch[k] = torch.tensor([f[k] for f in features])
/n/fs/nlp-pranjal/miniconda3/lib/python3.9/site-packages/transformers/data/data_collator.py:132: UserWarning: Creating a tensor from a list of numpy.ndarrays is extremely slow. Please consider converting the list to a single numpy.ndarray with numpy.array() before converting to a tensor. (Triggered internally at ../torch/csrc/utils/tensor_new.cpp:210.)
batch[k] = torch.tensor([f[k] for f in features])
/n/fs/nlp-pranjal/miniconda3/lib/python3.9/site-packages/transformers/data/data_collator.py:132: UserWarning: Creating a tensor from a list of numpy.ndarrays is extremely slow. Please consider converting the list to a single numpy.ndarray with numpy.array() before converting to a tensor. (Triggered internally at ../torch/csrc/utils/tensor_new.cpp:210.)
batch[k] = torch.tensor([f[k] for f in features])
/n/fs/nlp-pranjal/miniconda3/lib/python3.9/site-packages/transformers/data/data_collator.py:132: UserWarning: Creating a tensor from a list of numpy.ndarrays is extremely slow. Please consider converting the list to a single numpy.ndarray with numpy.array() before converting to a tensor. (Triggered internally at ../torch/csrc/utils/tensor_new.cpp:210.)
batch[k] = torch.tensor([f[k] for f in features])
/n/fs/nlp-pranjal/miniconda3/lib/python3.9/site-packages/transformers/data/data_collator.py:132: UserWarning: Creating a tensor from a list of numpy.ndarrays is extremely slow. Please consider converting the list to a single numpy.ndarray with numpy.array() before converting to a tensor. (Triggered internally at ../torch/csrc/utils/tensor_new.cpp:210.)
batch[k] = torch.tensor([f[k] for f in features])
/n/fs/nlp-pranjal/miniconda3/lib/python3.9/site-packages/transformers/data/data_collator.py:132: UserWarning: Creating a tensor from a list of numpy.ndarrays is extremely slow. Please consider converting the list to a single numpy.ndarray with numpy.array() before converting to a tensor. (Triggered internally at ../torch/csrc/utils/tensor_new.cpp:210.)
batch[k] = torch.tensor([f[k] for f in features])
/n/fs/nlp-pranjal/miniconda3/lib/python3.9/site-packages/transformers/data/data_collator.py:132: UserWarning: Creating a tensor from a list of numpy.ndarrays is extremely slow. Please consider converting the list to a single numpy.ndarray with numpy.array() before converting to a tensor. (Triggered internally at ../torch/csrc/utils/tensor_new.cpp:210.)
batch[k] = torch.tensor([f[k] for f in features])
/n/fs/nlp-pranjal/miniconda3/lib/python3.9/site-packages/transformers/data/data_collator.py:132: UserWarning: Creating a tensor from a list of numpy.ndarrays is extremely slow. Please consider converting the list to a single numpy.ndarray with numpy.array() before converting to a tensor. (Triggered internally at ../torch/csrc/utils/tensor_new.cpp:210.)
batch[k] = torch.tensor([f[k] for f in features])
/n/fs/nlp-pranjal/miniconda3/lib/python3.9/site-packages/transformers/data/data_collator.py:132: UserWarning: Creating a tensor from a list of numpy.ndarrays is extremely slow. Please consider converting the list to a single numpy.ndarray with numpy.array() before converting to a tensor. (Triggered internally at ../torch/csrc/utils/tensor_new.cpp:210.)
batch[k] = torch.tensor([f[k] for f in features])
/n/fs/nlp-pranjal/miniconda3/lib/python3.9/site-packages/transformers/data/data_collator.py:132: UserWarning: Creating a tensor from a list of numpy.ndarrays is extremely slow. Please consider converting the list to a single numpy.ndarray with numpy.array() before converting to a tensor. (Triggered internally at ../torch/csrc/utils/tensor_new.cpp:210.)
batch[k] = torch.tensor([f[k] for f in features])
/n/fs/nlp-pranjal/miniconda3/lib/python3.9/site-packages/transformers/data/data_collator.py:132: UserWarning: Creating a tensor from a list of numpy.ndarrays is extremely slow. Please consider converting the list to a single numpy.ndarray with numpy.array() before converting to a tensor. (Triggered internally at ../torch/csrc/utils/tensor_new.cpp:210.)
batch[k] = torch.tensor([f[k] for f in features])
/n/fs/nlp-pranjal/miniconda3/lib/python3.9/site-packages/transformers/data/data_collator.py:132: UserWarning: Creating a tensor from a list of numpy.ndarrays is extremely slow. Please consider converting the list to a single numpy.ndarray with numpy.array() before converting to a tensor. (Triggered internally at ../torch/csrc/utils/tensor_new.cpp:210.)
batch[k] = torch.tensor([f[k] for f in features])
/n/fs/nlp-pranjal/miniconda3/lib/python3.9/site-packages/transformers/data/data_collator.py:132: UserWarning: Creating a tensor from a list of numpy.ndarrays is extremely slow. Please consider converting the list to a single numpy.ndarray with numpy.array() before converting to a tensor. (Triggered internally at ../torch/csrc/utils/tensor_new.cpp:210.)
batch[k] = torch.tensor([f[k] for f in features])
/n/fs/nlp-pranjal/miniconda3/lib/python3.9/site-packages/transformers/data/data_collator.py:132: UserWarning: Creating a tensor from a list of numpy.ndarrays is extremely slow. Please consider converting the list to a single numpy.ndarray with numpy.array() before converting to a tensor. (Triggered internally at ../torch/csrc/utils/tensor_new.cpp:210.)
batch[k] = torch.tensor([f[k] for f in features])
Traceback (most recent call last):
File "/n/fs/nlp-pranjal/SemSup-LMLC/cleaned_code/main.py", line 598, in <module>
File "/n/fs/nlp-pranjal/SemSup-LMLC/cleaned_code/main.py", line 513, in main
data_args.max_train_samples if data_args.max_train_samples is not None else len(train_dataset)
File "/n/fs/nlp-pranjal/miniconda3/lib/python3.9/site-packages/transformers/trainer.py", line 1409, in train
return inner_training_loop(
File "/n/fs/nlp-pranjal/miniconda3/lib/python3.9/site-packages/transformers/trainer.py", line 1651, in _inner_training_loop
tr_loss_step = self.training_step(model, inputs)
File "/n/fs/nlp-pranjal/miniconda3/lib/python3.9/site-packages/transformers/trainer.py", line 2349, in training_step
loss = self.compute_loss(model, inputs)
File "/n/fs/nlp-pranjal/miniconda3/lib/python3.9/site-packages/transformers/trainer.py", line 2381, in compute_loss
outputs = model(**inputs)
File "/n/fs/nlp-pranjal/miniconda3/lib/python3.9/site-packages/torch/nn/modules/module.py", line 1110, in _call_impl
return forward_call(*input, **kwargs)
File "/n/fs/nlp-pranjal/miniconda3/lib/python3.9/site-packages/torch/nn/parallel/data_parallel.py", line 168, in forward
outputs = self.parallel_apply(replicas, inputs, kwargs)
File "/n/fs/nlp-pranjal/miniconda3/lib/python3.9/site-packages/torch/nn/parallel/data_parallel.py", line 178, in parallel_apply
return parallel_apply(replicas, inputs, kwargs, self.device_ids[:len(replicas)])
File "/n/fs/nlp-pranjal/miniconda3/lib/python3.9/site-packages/torch/nn/parallel/parallel_apply.py", line 86, in parallel_apply
output.reraise()
File "/n/fs/nlp-pranjal/miniconda3/lib/python3.9/site-packages/torch/_utils.py", line 457, in reraise
raise exception
RuntimeError: Caught RuntimeError in replica 0 on device 0.
Original Traceback (most recent call last):
File "/n/fs/nlp-pranjal/miniconda3/lib/python3.9/site-packages/torch/nn/parallel/parallel_apply.py", line 61, in _worker
output = module(*input, **kwargs)
File "/n/fs/nlp-pranjal/miniconda3/lib/python3.9/site-packages/torch/nn/modules/module.py", line 1110, in _call_impl
return forward_call(*input, **kwargs)
File "/n/fs/nlp-pranjal/SemSup-LMLC/cleaned_code/src/models.py", line 459, in forward
File "/n/fs/nlp-pranjal/SemSup-LMLC/cleaned_code/src/models.py", line 247, in coil_forward
lab_reps = self.tok_proj(outputs_lab.last_hidden_state @ self.label_projection.weight) # Q * LQ * d
File "/n/fs/nlp-pranjal/SemSup-LMLC/cleaned_code/src/models.py", line 399, in forward_label_embeddings
desc_attention_mask: Optional[List[int]] = None,
File "/n/fs/nlp-pranjal/miniconda3/lib/python3.9/site-packages/torch/nn/modules/module.py", line 1110, in _call_impl
return forward_call(*input, **kwargs)
File "/n/fs/nlp-pranjal/miniconda3/lib/python3.9/site-packages/transformers/models/bert/modeling_bert.py", line 1018, in forward
encoder_outputs = self.encoder(
File "/n/fs/nlp-pranjal/miniconda3/lib/python3.9/site-packages/torch/nn/modules/module.py", line 1110, in _call_impl
return forward_call(*input, **kwargs)
File "/n/fs/nlp-pranjal/miniconda3/lib/python3.9/site-packages/transformers/models/bert/modeling_bert.py", line 607, in forward
layer_outputs = layer_module(
File "/n/fs/nlp-pranjal/miniconda3/lib/python3.9/site-packages/torch/nn/modules/module.py", line 1110, in _call_impl
return forward_call(*input, **kwargs)
File "/n/fs/nlp-pranjal/miniconda3/lib/python3.9/site-packages/transformers/models/bert/modeling_bert.py", line 493, in forward
self_attention_outputs = self.attention(
File "/n/fs/nlp-pranjal/miniconda3/lib/python3.9/site-packages/torch/nn/modules/module.py", line 1110, in _call_impl
return forward_call(*input, **kwargs)
File "/n/fs/nlp-pranjal/miniconda3/lib/python3.9/site-packages/transformers/models/bert/modeling_bert.py", line 423, in forward
self_outputs = self.self(
File "/n/fs/nlp-pranjal/miniconda3/lib/python3.9/site-packages/torch/nn/modules/module.py", line 1110, in _call_impl
return forward_call(*input, **kwargs)
File "/n/fs/nlp-pranjal/miniconda3/lib/python3.9/site-packages/transformers/models/bert/modeling_bert.py", line 355, in forward
attention_probs = self.dropout(attention_probs)
File "/n/fs/nlp-pranjal/miniconda3/lib/python3.9/site-packages/torch/nn/modules/module.py", line 1110, in _call_impl
return forward_call(*input, **kwargs)
File "/n/fs/nlp-pranjal/miniconda3/lib/python3.9/site-packages/torch/nn/modules/dropout.py", line 58, in forward
return F.dropout(input, self.p, self.training, self.inplace)
File "/n/fs/nlp-pranjal/miniconda3/lib/python3.9/site-packages/torch/nn/functional.py", line 1279, in dropout
return _VF.dropout_(input, p, training) if inplace else _VF.dropout(input, p, training)
RuntimeError: CUDA out of memory. Tried to allocate 782.00 MiB (GPU 0; 10.76 GiB total capacity; 3.28 GiB already allocated; 61.69 MiB free; 3.65 GiB reserved in total by PyTorch) If reserved memory is >> allocated memory try setting max_split_size_mb to avoid fragmentation. See documentation for Memory Management and PYTORCH_CUDA_ALLOC_CONF
[31m╭─────────────────────────────── [39m[1mTraceback (most recent call last)[31m[22m ────────────────────────────────╮
[31m│[39m /n/fs/nlp-pranjal/SemSup-LMLC/cleaned_code/[1mmain.py[22m:[94m598[39m in [92m<module>[39m [31m│
[31m│[39m [31m│
[31m│[39m 595 │ main() [31m│
[31m│[39m [31m│
[31m│[39m /n/fs/nlp-pranjal/SemSup-LMLC/cleaned_code/[1mmain.py[22m:[94m513[39m in [92mmain[39m [31m│
[31m│[39m [31m│
[31m│[39m 510 │ │ train_result = trainer.train(resume_from_checkpoint=checkpoint) [31m│
[31m│[39m 511 │ │ metrics = train_result.metrics [31m│
[31m│[39m 512 │ │ max_train_samples = ( [31m│
[31m│[39m [31m❱ [39m513 │ │ │ data_args.max_train_samples [94mif[39m data_args.max_train_samples [95mis[39m [95mnot[39m [94mNone[39m [94melse[39m [31m│
[31m│[39m 514 │ │ ) [31m│
[31m│[39m 515 │ │ metrics[[33m"train_samples"[39m] = [96mmin[39m(max_train_samples, [96mlen[39m(train_dataset)) [31m│
[31m│[39m 516 [31m│
[31m│[39m [31m│
[31m│[39m /n/fs/nlp-pranjal/miniconda3/lib/python3.9/site-packages/transformers/[1mtrainer.py[22m:[94m1409[39m in [92mtrain[39m [31m│
[31m│[39m [31m│
[31m│[39m 1406 │ │ inner_training_loop = find_executable_batch_size( [31m│
[31m│[39m 1407 │ │ │ [96mself[39m._inner_training_loop, [96mself[39m._train_batch_size, args.auto_find_batch_size [31m│
[31m│[39m 1408 │ │ ) [31m│
[31m│[39m [31m❱ [39m1409 │ │ [94mreturn[39m inner_training_loop( [31m│
[31m│[39m 1410 │ │ │ args=args, [31m│
[31m│[39m 1411 │ │ │ resume_from_checkpoint=resume_from_checkpoint, [31m│
[31m│[39m 1412 │ │ │ trial=trial, [31m│
[31m│[39m [31m│
[31m│[39m /n/fs/nlp-pranjal/miniconda3/lib/python3.9/site-packages/transformers/[1mtrainer.py[22m:[94m1651[39m in [31m│
[31m│[39m [92m_inner_training_loop[39m [31m│
[31m│[39m [31m│
[31m│[39m 1648 │ │ │ │ │ [94mwith[39m model.no_sync(): [31m│
[31m│[39m 1649 │ │ │ │ │ │ tr_loss_step = [96mself[39m.training_step(model, inputs) [31m│
[31m│[39m 1650 │ │ │ │ [94melse[39m: [31m│
[31m│[39m [31m❱ [39m1651 │ │ │ │ │ tr_loss_step = [96mself[39m.training_step(model, inputs) [31m│
[31m│[39m 1652 │ │ │ │ [31m│
[31m│[39m 1653 │ │ │ │ [94mif[39m ( [31m│
[31m│[39m 1654 │ │ │ │ │ args.logging_nan_inf_filter [31m│
[31m│[39m [31m│
[31m│[39m /n/fs/nlp-pranjal/miniconda3/lib/python3.9/site-packages/transformers/[1mtrainer.py[22m:[94m2349[39m in [31m│
[31m│[39m [92mtraining_step[39m [31m│
[31m│[39m [31m│
[31m│[39m 2346 │ │ │ [94mreturn[39m loss_mb.reduce_mean().detach().to([96mself[39m.args.device) [31m│
[31m│[39m 2347 │ │ [31m│
[31m│[39m 2348 │ │ [94mwith[39m [96mself[39m.compute_loss_context_manager(): [31m│
[31m│[39m [31m❱ [39m2349 │ │ │ loss = [96mself[39m.compute_loss(model, inputs) [31m│
[31m│[39m 2350 │ │ [31m│
[31m│[39m 2351 │ │ [94mif[39m [96mself[39m.args.n_gpu > [94m1[39m: [31m│
[31m│[39m 2352 │ │ │ loss = loss.mean() # mean() to average on multi-gpu parallel training [31m│
[31m│[39m [31m│
[31m│[39m /n/fs/nlp-pranjal/miniconda3/lib/python3.9/site-packages/transformers/[1mtrainer.py[22m:[94m2381[39m in [31m│
[31m│[39m [92mcompute_loss[39m [31m│
[31m│[39m [31m│
[31m│[39m 2378 │ │ │ labels = inputs.pop([33m"labels"[39m) [31m│
[31m│[39m 2379 │ │ [94melse[39m: [31m│
[31m│[39m 2380 │ │ │ labels = [94mNone[39m [31m│
[31m│[39m [31m❱ [39m2381 │ │ outputs = model(**inputs) [31m│
[31m│[39m 2382 │ │ # Save past state if it exists [31m│
[31m│[39m 2383 │ │ # TODO: this needs to be fixed and made cleaner later. [31m│
[31m│[39m 2384 │ │ [94mif[39m [96mself[39m.args.past_index >= [94m0[39m: [31m│
[31m│[39m [31m│
[31m│[39m /n/fs/nlp-pranjal/miniconda3/lib/python3.9/site-packages/torch/nn/modules/[1mmodule.py[22m:[94m1110[39m in [31m│
[31m│[39m [92m_call_impl[39m [31m│
[31m│[39m [31m│
[31m│[39m 1107 │ │ # this function, and just call forward. [31m│
[31m│[39m 1108 │ │ [94mif[39m [95mnot[39m ([96mself[39m._backward_hooks [95mor[39m [96mself[39m._forward_hooks [95mor[39m [96mself[39m._forward_pre_hooks [95mo[39m [31m│
[31m│[39m 1109 │ │ │ │ [95mor[39m _global_forward_hooks [95mor[39m _global_forward_pre_hooks): [31m│
[31m│[39m [31m❱ [39m1110 │ │ │ [94mreturn[39m forward_call(*[96minput[39m, **kwargs) [31m│
[31m│[39m 1111 │ │ # Do not call functions when jit is used [31m│
[31m│[39m 1112 │ │ full_backward_hooks, non_full_backward_hooks = [], [] [31m│
[31m│[39m 1113 │ │ [94mif[39m [96mself[39m._backward_hooks [95mor[39m _global_backward_hooks: [31m│
[31m│[39m [31m│
[31m│[39m /n/fs/nlp-pranjal/miniconda3/lib/python3.9/site-packages/torch/nn/parallel/[1mdata_parallel.py[22m:[94m168[39m [31m│
[31m│[39m in [92mforward[39m [31m│
[31m│[39m [31m│
[31m│[39m 165 │ │ │ [94mif[39m [96mlen[39m([96mself[39m.device_ids) == [94m1[39m: [31m│
[31m│[39m 166 │ │ │ │ [94mreturn[39m [96mself[39m.module(*inputs[[94m0[39m], **kwargs[[94m0[39m]) [31m│
[31m│[39m 167 │ │ │ replicas = [96mself[39m.replicate([96mself[39m.module, [96mself[39m.device_ids[:[96mlen[39m(inputs)]) [31m│
[31m│[39m [31m❱ [39m168 │ │ │ outputs = [96mself[39m.parallel_apply(replicas, inputs, kwargs) [31m│
[31m│[39m 169 │ │ │ [94mreturn[39m [96mself[39m.gather(outputs, [96mself[39m.output_device) [31m│
[31m│[39m 170 │ [31m│
[31m│[39m 171 │ [94mdef[39m [92mreplicate[39m([96mself[39m, module, device_ids): [31m│
[31m│[39m [31m│
[31m│[39m /n/fs/nlp-pranjal/miniconda3/lib/python3.9/site-packages/torch/nn/parallel/[1mdata_parallel.py[22m:[94m178[39m [31m│
[31m│[39m in [92mparallel_apply[39m [31m│
[31m│[39m [31m│
[31m│[39m 175 │ │ [94mreturn[39m scatter_kwargs(inputs, kwargs, device_ids, dim=[96mself[39m.dim) [31m│
[31m│[39m 176 │ [31m│
[31m│[39m 177 │ [94mdef[39m [92mparallel_apply[39m([96mself[39m, replicas, inputs, kwargs): [31m│
[31m│[39m [31m❱ [39m178 │ │ [94mreturn[39m parallel_apply(replicas, inputs, kwargs, [96mself[39m.device_ids[:[96mlen[39m(replicas)]) [31m│
[31m│[39m 179 │ [31m│
[31m│[39m 180 │ [94mdef[39m [92mgather[39m([96mself[39m, outputs, output_device): [31m│
[31m│[39m 181 │ │ [94mreturn[39m gather(outputs, output_device, dim=[96mself[39m.dim) [31m│
[31m│[39m [31m│
[31m│[39m /n/fs/nlp-pranjal/miniconda3/lib/python3.9/site-packages/torch/nn/parallel/[1mparallel_apply.py[22m:[94m86[39m [31m│
[31m│[39m in [92mparallel_apply[39m [31m│
[31m│[39m [31m│
[31m│[39m 83 │ [94mfor[39m i [95min[39m [96mrange[39m([96mlen[39m(inputs)): [31m│
[31m│[39m 84 │ │ output = results[i] [31m│
[31m│[39m 85 │ │ [94mif[39m [96misinstance[39m(output, ExceptionWrapper): [31m│
[31m│[39m [31m❱ [39m86 │ │ │ output.reraise() [31m│
[31m│[39m 87 │ │ outputs.append(output) [31m│
[31m│[39m 88 │ [94mreturn[39m outputs [31m│
[31m│[39m 89 [31m│
[31m│[39m [31m│
[31m│[39m /n/fs/nlp-pranjal/miniconda3/lib/python3.9/site-packages/torch/[1m_utils.py[22m:[94m457[39m in [92mreraise[39m [31m│
[31m│[39m [31m│
[31m│[39m 454 │ │ │ # If the exception takes multiple arguments, don't try to [31m│
[31m│[39m 455 │ │ │ # instantiate since we don't know how to [31m│
[31m│[39m 456 │ │ │ [94mraise[39m [96mRuntimeError[39m(msg) [94mfrom[39m [96mNone[39m [31m│
[31m│[39m [31m❱ [39m457 │ │ [94mraise[39m exception [31m│
[31m│[39m 458 [31m│
[31m│[39m 459 [31m│
[31m│[39m 460 [94mdef[39m [92m_get_available_device_type[39m(): [31m│
[31m╰──────────────────────────────────────────────────────────────────────────────────────────────────╯
[1mRuntimeError: [22mCaught RuntimeError in replica [1m0[22m on device [1m0[22m.
Original Traceback [1m([22mmost recent call last[1m)[22m:
File [32m"/n/fs/nlp-pranjal/miniconda3/lib/python3.9/site-packages/torch/nn/parallel/parallel_apply.py"[39m, line [1m61[22m, in _worker
output = [1mmodule([22m*input, **kwargs[1m)
File [32m"/n/fs/nlp-pranjal/miniconda3/lib/python3.9/site-packages/torch/nn/modules/module.py"[39m, line [1m1110[22m, in _call_impl
return [1mforward_call([22m*input, **kwargs[1m)
File [32m"/n/fs/nlp-pranjal/SemSup-LMLC/cleaned_code/src/models.py"[39m, line [1m459[22m, in forward
File [32m"/n/fs/nlp-pranjal/SemSup-LMLC/cleaned_code/src/models.py"[39m, line [1m247[22m, in coil_forward
lab_reps = [1mself.tok_proj([22moutputs_lab.last_hidden_state @ self.label_projection.weight[1m)[22m # Q * LQ * d
File [32m"/n/fs/nlp-pranjal/SemSup-LMLC/cleaned_code/src/models.py"[39m, line [1m399[22m, in forward_label_embeddings
desc_attention_mask: Optional[1m[[22mList[1m[[22mint[1m]][22m = [3mNone[23m,
File [32m"/n/fs/nlp-pranjal/miniconda3/lib/python3.9/site-packages/torch/nn/modules/module.py"[39m, line [1m1110[22m, in _call_impl
return [1mforward_call([22m*input, **kwargs[1m)
File [32m"/n/fs/nlp-pranjal/miniconda3/lib/python3.9/site-packages/transformers/models/bert/modeling_bert.py"[39m, line [1m1018[22m, in forward
encoder_outputs = [1mself.encoder(
File [32m"/n/fs/nlp-pranjal/miniconda3/lib/python3.9/site-packages/torch/nn/modules/module.py"[39m, line [1m1110[22m, in _call_impl
return [1mforward_call([22m*input, **kwargs[1m)
File [32m"/n/fs/nlp-pranjal/miniconda3/lib/python3.9/site-packages/transformers/models/bert/modeling_bert.py"[39m, line [1m607[22m, in forward
layer_outputs = [1mlayer_module(
File [32m"/n/fs/nlp-pranjal/miniconda3/lib/python3.9/site-packages/torch/nn/modules/module.py"[39m, line [1m1110[22m, in _call_impl
return [1mforward_call([22m*input, **kwargs[1m)
File [32m"/n/fs/nlp-pranjal/miniconda3/lib/python3.9/site-packages/transformers/models/bert/modeling_bert.py"[39m, line [1m493[22m, in forward
self_attention_outputs = [1mself.attention(
File [32m"/n/fs/nlp-pranjal/miniconda3/lib/python3.9/site-packages/torch/nn/modules/module.py"[39m, line [1m1110[22m, in _call_impl
return [1mforward_call([22m*input, **kwargs[1m)
File [32m"/n/fs/nlp-pranjal/miniconda3/lib/python3.9/site-packages/transformers/models/bert/modeling_bert.py"[39m, line [1m423[22m, in forward
self_outputs = [1mself.self(
File [32m"/n/fs/nlp-pranjal/miniconda3/lib/python3.9/site-packages/torch/nn/modules/module.py"[39m, line [1m1110[22m, in _call_impl
return [1mforward_call([22m*input, **kwargs[1m)
File [32m"/n/fs/nlp-pranjal/miniconda3/lib/python3.9/site-packages/transformers/models/bert/modeling_bert.py"[39m, line [1m355[22m, in forward
attention_probs = [1mself.dropout([22mattention_probs[1m)
File [32m"/n/fs/nlp-pranjal/miniconda3/lib/python3.9/site-packages/torch/nn/modules/module.py"[39m, line [1m1110[22m, in _call_impl
return [1mforward_call([22m*input, **kwargs[1m)
File [32m"/n/fs/nlp-pranjal/miniconda3/lib/python3.9/site-packages/torch/nn/modules/dropout.py"[39m, line [1m58[22m, in forward
return [1mF.dropout([22minput, self.p, self.training, self.inplace[1m)
File [32m"/n/fs/nlp-pranjal/miniconda3/lib/python3.9/site-packages/torch/nn/functional.py"[39m, line [1m1279[22m, in dropout
return [1m_VF.dropout_([22minput, p, training[1m)[22m if inplace else [1m_VF.dropout([22minput, p, training[1m)
RuntimeError: CUDA out of memory. Tried to allocate [1m782.00[22m MiB [1m([22mGPU [1m0[22m; [1m10.76[22m GiB total capacity; [1m3.28[22m GiB already allocated; [1m61.69[22m MiB free; [1m3.65[22m GiB reserved in total
by PyTorch[1m)[22m If reserved memory is >> allocated memory try setting max_split_size_mb to avoid fragmentation. See documentation for Memory Management and
PYTORCH_CUDA_ALLOC_CONF |