Views
No views yet
| Name | Quant method | Size |
|---|---|---|
| Neural-Mistral-7B.Q2_K.gguf | Q2_K | 2.53GB |
| Neural-Mistral-7B.IQ3_XS.gguf | IQ3_XS | 2.81GB |
| Neural-Mistral-7B.IQ3_S.gguf | IQ3_S | 2.96GB |
| Neural-Mistral-7B.Q3_K_S.gguf | Q3_K_S | 2.95GB |
| Neural-Mistral-7B.IQ3_M.gguf | IQ3_M | 3.06GB |
| Neural-Mistral-7B.Q3_K.gguf | Q3_K | 3.28GB |
| Neural-Mistral-7B.Q3_K_M.gguf | Q3_K_M | 3.28GB |
| Neural-Mistral-7B.Q3_K_L.gguf | Q3_K_L | 3.56GB |
| Neural-Mistral-7B.IQ4_XS.gguf | IQ4_XS | 3.67GB |
| Neural-Mistral-7B.Q4_0.gguf | Q4_0 | 3.83GB |
| Neural-Mistral-7B.IQ4_NL.gguf | IQ4_NL | 3.87GB |
| Neural-Mistral-7B.Q4_K_S.gguf | Q4_K_S | 3.86GB |
| Neural-Mistral-7B.Q4_K.gguf | Q4_K | 4.07GB |
| Neural-Mistral-7B.Q4_K_M.gguf | Q4_K_M | 4.07GB |
| Neural-Mistral-7B.Q4_1.gguf | Q4_1 | 4.24GB |
| Neural-Mistral-7B.Q5_0.gguf | Q5_0 | 4.65GB |
| Neural-Mistral-7B.Q5_K_S.gguf | Q5_K_S | 4.65GB |
| Neural-Mistral-7B.Q5_K.gguf | Q5_K | 4.78GB |
| Neural-Mistral-7B.Q5_K_M.gguf | Q5_K_M | 4.78GB |
| Neural-Mistral-7B.Q5_1.gguf | Q5_1 | 5.07GB |
| Neural-Mistral-7B.Q6_K.gguf | Q6_K | 5.53GB |
| Neural-Mistral-7B.Q8_0.gguf | Q8_0 | 7.17GB |
[INST] and [/INST] tokens. The very first instruction should begin with a begin of sentence id. The next instructions should not. The assistant generation will be ended by the end-of-sentence token id.text = "<s>[INST] What is your favourite condiment? [/INST]"
"Well, I'm quite partial to a good squeeze of fresh lemon juice. It adds just the right amount of zesty flavour to whatever I'm cooking up in the kitchen!</s> "
"[INST] Do you have mayonnaise recipes? [/INST]"apply_chat_template() method:1from transformers import AutoModelForCausalLM, AutoTokenizer
2
3device = "cuda" # the device to load the model onto
4
5model = AutoModelForCausalLM.from_pretrained("mistralai/Mistral-7B-Instruct-v0.2")
6tokenizer = AutoTokenizer.from_pretrained("mistralai/Mistral-7B-Instruct-v0.2")
7
8messages = [
9 {"role": "user", "content": "What is your favourite condiment?"},
10 {"role": "assistant", "content": "Well, I'm quite partial to a good squeeze of fresh lemon juice. It adds just the right amount of zesty flavour to whatever I'm cooking up in the kitchen!"},
11 {"role": "user", "content": "Do you have mayonnaise recipes?"}
12]
13
14encodeds = tokenizer.apply_chat_template(messages, return_tensors="pt")
15
16model_inputs = encodeds.to(device)
17model.to(device)
18
19generated_ids = model.generate(model_inputs, max_new_tokens=1000, do_sample=True)
20decoded = tokenizer.batch_decode(generated_ids)
21print(decoded[0])# Format chosen answer
chosen = example['chosen'] + tokenizer.eos_token
# Format rejected answer
rejected = example['rejected'] + tokenizer.eos_token
return {
"prompt": prompt,
"chosen": chosen,
"rejected": rejected,
}