self.inv_freq[None, :, None].float().expand(position_ids.shape[0], -1, 1) AttributeError: ‘dict’ object has no attribute ‘shape’` when I use llama 2
Traceback (most recent call last): File "/home/songbaiyang/work/tvl_llama/main_pretrain.py", line 253, in <module> main(args) File "/home/songbaiyang/work/tvl_llama/main_pretrain.py", line 217, in main train_stats = train_one_epoch( File "/home/songbaiyang/work/tvl_llama/engine_pretrain.py", line 46, in train_one_epoch c_loss, m_loss = model(examples, labels, observations) File "/home/songbaiyang/.conda/envs/my_tvl/lib/python3.10/site-packages/torch/nn/modules/module.py", line 1518, in _wrapped_call_impl return self._call_impl(*args, **kwargs) File "/home/songbaiyang/.conda/envs/my_tvl/lib/python3.10/site-packages/torch/nn/modules/module.py", line 1527, in _call_impl return forward_call(*args, **kwargs) File "/home/songbaiyang/.conda/envs/my_tvl/lib/python3.10/site-packages/accelerate/hooks.py", line 166, in new_forward output = module._old_forward(*args, **kwargs) File "/home/songbaiyang/.conda/envs/my_tvl/lib/python3.10/site-packages/transformers/models/llama/modeling_llama.py", line 1208, in forward outputs = self.model( File "/home/songbaiyang/.conda/envs/my_tvl/lib/python3.10/site-packages/torch/nn/modules/module.py", line 1518, in _wrapped_call_impl return self._call_impl(*args, **kwargs) File "/home/songbaiyang/.conda/envs/my_tvl/lib/python3.10/site-packages/torch/nn/modules/module.py", line 1527, in _call_impl return forward_call(*args, **kwargs) File "/home/songbaiyang/.conda/envs/my_tvl/lib/python3.10/site-packages/transformers/models/llama/modeling_llama.py", line 1018, in forward layer_outputs = decoder_layer( File "/home/songbaiyang/.conda/envs/my_tvl/lib/python3.10/site-packages/torch/nn/modules/module.py", line 1518, in _wrapped_call_impl return self._call_impl(*args, **kwargs) File "/home/songbaiyang/.conda/envs/my_tvl/lib/python3.10/site-packages/torch/nn/modules/module.py", line 1527, in _call_impl return forward_call(*args, **kwargs) File "/home/songbaiyang/.conda/envs/my_tvl/lib/python3.10/site-packages/accelerate/hooks.py", line 166, in new_forward output = module._old_forward(*args, **kwargs) File "/home/songbaiyang/.conda/envs/my_tvl/lib/python3.10/site-packages/transformers/models/llama/modeling_llama.py", line 741, in forward hidden_states, self_attn_weights, present_key_value = self.self_attn( File "/home/songbaiyang/.conda/envs/my_tvl/lib/python3.10/site-packages/torch/nn/modules/module.py", line 1518, in _wrapped_call_impl return self._call_impl(*args, **kwargs) File "/home/songbaiyang/.conda/envs/my_tvl/lib/python3.10/site-packages/torch/nn/modules/module.py", line 1527, in _call_impl return forward_call(*args, **kwargs) File "/home/songbaiyang/.conda/envs/my_tvl/lib/python3.10/site-packages/accelerate/hooks.py", line 166, in new_forward output = module._old_forward(*args, **kwargs) File "/home/songbaiyang/.conda/envs/my_tvl/lib/python3.10/site-packages/transformers/models/llama/modeling_llama.py", line 644, in forward cos, sin = self.rotary_emb(value_states, position_ids) File "/home/songbaiyang/.conda/envs/my_tvl/lib/python3.10/site-packages/torch/nn/modules/module.py", line 1518, in _wrapped_call_impl return self._call_impl(*args, **kwargs) File "/home/songbaiyang/.conda/envs/my_tvl/lib/python3.10/site-packages/torch/nn/modules/module.py", line 1527, in _call_impl return forward_call(*args, **kwargs) File "/home/songbaiyang/.conda/envs/my_tvl/lib/python3.10/site-packages/torch/utils/_contextlib.py", line 115, in decorate_context return func(*args, **kwargs) File "/home/songbaiyang/.conda/envs/my_tvl/lib/python3.10/site-packages/transformers/models/llama/modeling_llama.py", line 134, in forward inv_freq_expanded = self.inv_freq[None, :, None].float().expand(position_ids.shape[0], -1, 1) AttributeError: 'dict' object has no attribute 'shape'
How to clear the ‘previous’ cache when I’m using use_cache=True option in model.generate()
issue about KV cache in streaming(generation) (huggingface transformers)
How to clear the ‘old’ cache when I’m using use_cache=True option in model.generate()
issue about KV cache in streaming(generation) (huggingface transformers)
How to clear the cache when I’m using use_cache=True option in model.generate()
issue about KV cache in streaming(generation) (huggingface transformers)