-
`tokenizer = AutoTokenizer.from_pretrained(checkpoint)` as defined here - https://github.com/bigcode-project/starcoder#code-generation
pulls 7 checkpoint files, ~9GB each. Is this the intended behavi…
-
Just want to figure out the reasons:
- not support C++?
- model performance too poor?
- other configuration mistake?
## appendix
- C++ case:
 for questions
- [X] I'm not able to find an [open issue](ht…
-
In ablation study, you show multi round revision works better than single round revision. You also show wizardcoder outperforms Pretrained model starcoder. However, it is unknown how much wizardcoder …
-
When I use the infer command(inference_wizardcoder.py), an error occurs
python src/inference_wizardcoder.py --base_model "bigcode/starcoder" --input_data_path "./input.jsonl" --output_data_…
-
It’s important to avoid ever using claims that are dependent on thousands of variables changing each day, such as, “With a context length of over 8,000 tokens, the StarCoder models can process more in…
-
There is currently only
- baichat
- catgpt
- huggingchat
- openaigpt35turbo
enabled by default.
But there is
- alpacalora
- baichat
- bard (disabled for now)
- catgpt
- hfdialogpt
- hfg…
-
Hi all, I've set up Starcoder as follows:
```
gen_checkpoint = "bigcode/starcoder"
gen_device = "cuda"
gen_tokenizer, gen_model = setup_model_tokenizer(
gen_checkpoint, bit_4=False, device=…
-
Looks like GPU usage almost doubles during saving (save_pretrained - get_peft_model_state_dict function). Is there a way to avoid this?
stack trace:
```Traceback (most recent call last):
Fil…