Downloads · 30 days
126
9% of all-time downloads
sawao_-_gemma2-2b-sft-gguf
RichardErkhov/sawao_-_gemma2-2b-sft-gguf
sawao_-_gemma2-2b-sft-gguf is a machine learning model from RichardErkhov. Use it for the machine learning task on the model card, and read the license before you ship it in a product.
All-time downloads
1.3K
Public
Repo size
32.3 GB
Likes
0
Public
Repo makeup
Click a slice to open those files.
.gguf37.3 GB · 100%
At a glance
- Access
- Public
- Created
- Feb 26, 2025
- Updated
- Feb 26, 2025
- SHA
- 69f8d76a
Model card
From the Hugging Face model README
Quantization made by Richard Erkhov.
gemma2-2b-sft - GGUF
- Model creator: https://huggingface.co/sawao/
- Original model: https://huggingface.co/sawao/gemma2-2b-sft/
| Name | Quant method | Size |
|---|---|---|
| gemma2-2b-sft.Q2_K.gguf | Q2_K | 1.15GB |
| gemma2-2b-sft.IQ3_XS.gguf | IQ3_XS | 1.22GB |
| gemma2-2b-sft.IQ3_S.gguf | IQ3_S | 1.27GB |
| gemma2-2b-sft.Q3_K_S.gguf | Q3_K_S | 1.27GB |
| gemma2-2b-sft.IQ3_M.gguf | IQ3_M | 1.3GB |
| gemma2-2b-sft.Q3_K.gguf | Q3_K | 1.36GB |
| gemma2-2b-sft.Q3_K_M.gguf | Q3_K_M | 1.36GB |
| gemma2-2b-sft.Q3_K_L.gguf | Q3_K_L | 1.44GB |
| gemma2-2b-sft.IQ4_XS.gguf | IQ4_XS | 1.47GB |
| gemma2-2b-sft.Q4_0.gguf | Q4_0 | 1.52GB |
| gemma2-2b-sft.IQ4_NL.gguf | IQ4_NL | 1.53GB |
| gemma2-2b-sft.Q4_K_S.gguf | Q4_K_S | 1.53GB |
| gemma2-2b-sft.Q4_K.gguf | Q4_K | 1.59GB |
| gemma2-2b-sft.Q4_K_M.gguf | Q4_K_M | 1.59GB |
| gemma2-2b-sft.Q4_1.gguf | Q4_1 | 1.64GB |
| gemma2-2b-sft.Q5_0.gguf | Q5_0 | 1.75GB |
| gemma2-2b-sft.Q5_K_S.gguf | Q5_K_S | 1.75GB |
| gemma2-2b-sft.Q5_K.gguf | Q5_K | 1.79GB |
| gemma2-2b-sft.Q5_K_M.gguf | Q5_K_M | 1.79GB |
| gemma2-2b-sft.Q5_1.gguf | Q5_1 | 1.87GB |
| gemma2-2b-sft.Q6_K.gguf | Q6_K | 2.0GB |
| gemma2-2b-sft.Q8_0.gguf | Q8_0 | 2.59GB |
Original model description:
library_name: transformers tags: []
gemma2-2b-sft
base_model: gemma2_2b gemma2 27bにより生成したself instructデータ10kによりinstrution tuningを実施
Eval
elyza-task-100
Use
model = AutoModelForCausalLM.from_pretrained(
model_name,
torch_dtype=torch.bfloat16,
device_map="auto" # GPU自動割り当て
)
tokenizer = AutoTokenizer.from_pretrained(model_name)
messages_list = [
[
{"role": "user", "content": "仕事の熱意を取り戻すためのアイデアを5つ挙げてください。"}
]
]
prompts = [self.tokenizer.apply_chat_template(messages, tokenize=False, add_generation_prompt=True) for messages in messages_list]
inputs = self.tokenizer(prompts, return_tensors="pt", padding=True).to(self.model.device)
outputs = self.model.generate(
**inputs,
temperature=self.generate_configs["temperature"],
max_new_tokens=self.generate_configs["max_new_tokens"],
top_p=self.generate_configs["top_p"],
top_k=self.generate_configs["top_k"],
repetition_penalty=self.generate_configs["repetition_penalty"],
pad_token_id=self.tokenizer.pad_token_id,
eos_token_id=self.tokenizer.eos_token_id,
)
print(outputs)
- Created
- Feb 26, 2025
- Updated
- Feb 26, 2025