Support T5 models #643
Description
I tried to load a T5 model but it seems not supported.
---------------------------------------------------------------------------
NotImplementedError Traceback (most recent call last)
Cell In[5], line 7
4 dtype = None # None for auto detection. Float16 for Tesla T4, V100, Bfloat16 for Ampere+
5 load_in_4bit = True # Use 4bit quantization to reduce memory usage. Can be False.
----> 7 model, tokenizer = FastLanguageModel.from_pretrained(
8 model_name = "google-t5/t5-large", # Choose ANY! eg mistralai/Mistral-7B-Instruct-v0.2
9 max_seq_length = max_seq_length,
10 dtype = dtype,
11 load_in_4bit = load_in_4bit,
12 # token = "hf_...", # use one if using gated models like meta-llama/Llama-2-7b-hf
13 )
File ~/anaconda3/envs/pytorch_p310/lib/python3.10/site-packages/unsloth/models/loader.py:127, in FastLanguageModel.from_pretrained(model_name, max_seq_length, dtype, load_in_4bit, token, device_map, rope_scaling, fix_tokenizer, trust_remote_code, use_gradient_checkpointing, resize_model_vocab, revision, *args, **kwargs)
125 dispatch_model = FastQwen2Model
126 else:
--> 127 raise NotImplementedError(
128 f"Unsloth: {model_name} not supported yet!\n"\
129 "Make an issue to https://github.com/unslothai/unsloth!",
130 )
131 pass
133 # Check if this is local model since the tokenizer gets overwritten
NotImplementedError: Unsloth: google-t5/t5-large not supported yet!
Make an issue to https://github.com/unslothai/unsloth!