Downloads · 30 days
50
2% of all-time downloads
RichardErkhov/piotr25691_-_thea-3b-25r-gguf
piotr25691_-_thea-3b-25r-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.
Downloads · 30 days
50
2% of all-time downloads
All-time downloads
2.2K
Public
Repo size
38.1 GB
Likes
0
Public
Click a slice to open those files.
.gguf44.2 GB · 100%
From the Hugging Face model README
Quantization made by Richard Erkhov.
thea-3b-25r - GGUF
| Name | Quant method | Size |
|---|---|---|
| thea-3b-25r.Q2_K.gguf | Q2_K | 1.27GB |
| thea-3b-25r.IQ3_XS.gguf | IQ3_XS | 1.38GB |
| thea-3b-25r.IQ3_S.gguf | IQ3_S | 1.44GB |
| thea-3b-25r.Q3_K_S.gguf | Q3_K_S | 1.44GB |
| thea-3b-25r.IQ3_M.gguf | IQ3_M | 1.49GB |
| thea-3b-25r.Q3_K.gguf | Q3_K | 1.57GB |
| thea-3b-25r.Q3_K_M.gguf | Q3_K_M | 1.57GB |
| thea-3b-25r.Q3_K_L.gguf | Q3_K_L | 1.69GB |
| thea-3b-25r.IQ4_XS.gguf | IQ4_XS | 1.71GB |
| thea-3b-25r.Q4_0.gguf | Q4_0 | 1.79GB |
| thea-3b-25r.IQ4_NL.gguf | IQ4_NL | 1.79GB |
| thea-3b-25r.Q4_K_S.gguf | Q4_K_S | 1.8GB |
| thea-3b-25r.Q4_K.gguf | Q4_K | 1.88GB |
| thea-3b-25r.Q4_K_M.gguf | Q4_K_M | 1.88GB |
| thea-3b-25r.Q4_1.gguf | Q4_1 | 1.95GB |
| thea-3b-25r.Q5_0.gguf | Q5_0 | 2.11GB |
| thea-3b-25r.Q5_K_S.gguf | Q5_K_S | 2.11GB |
| thea-3b-25r.Q5_K.gguf | Q5_K | 2.16GB |
| thea-3b-25r.Q5_K_M.gguf | Q5_K_M | 2.16GB |
| thea-3b-25r.Q5_1.gguf | Q5_1 | 2.28GB |
| thea-3b-25r.Q6_K.gguf | Q6_K | 2.46GB |
| thea-3b-25r.Q8_0.gguf | Q8_0 | 3.19GB |
language:
An uncensored reasoning Llama 3.2 3B model trained on reasoning data.
It has been trained using improved training code, and gives an improved performance. Here is what inference code you should use:
from transformers import AutoModelForCausalLM, AutoTokenizer
MAX_REASONING_TOKENS = 1024
MAX_RESPONSE_TOKENS = 512
model_name = "piotr25691/thea-3b-25r"
model = AutoModelForCausalLM.from_pretrained(model_name, torch_dtype="auto", device_map="auto")
tokenizer = AutoTokenizer.from_pretrained(model_name)
prompt = "Which is greater 9.9 or 9.11 ??"
messages = [
{"role": "user", "content": prompt}
]
# Generate reasoning
reasoning_template = tokenizer.apply_chat_template(messages, tokenize=False, add_reasoning_prompt=True)
reasoning_inputs = tokenizer(reasoning_template, return_tensors="pt").to(model.device)
reasoning_ids = model.generate(**reasoning_inputs, max_new_tokens=MAX_REASONING_TOKENS)
reasoning_output = tokenizer.decode(reasoning_ids[0, reasoning_inputs.input_ids.shape[1]:], skip_special_tokens=True)
print("REASONING: " + reasoning_output)
# Generate answer
messages.append({"role": "reasoning", "content": reasoning_output})
response_template = tokenizer.apply_chat_template(messages, tokenize=False, add_generation_prompt=True)
response_inputs = tokenizer(response_template, return_tensors="pt").to(model.device)
response_ids = model.generate(**response_inputs, max_new_tokens=MAX_RESPONSE_TOKENS)
response_output = tokenizer.decode(response_ids[0, response_inputs.input_ids.shape[1]:], skip_special_tokens=True)
print("ANSWER: " + response_output)
This Llama model was trained faster than Unsloth using custom training code.
Visit https://www.kaggle.com/code/piotr25691/distributed-llama-training-with-2xt4 to find out how you can finetune your models using BOTH of the Kaggle provided GPUs.
Additional thanks to @nicoboss for giving me access to his private supercomputer, enabling me to provide many more quants, at much higher speed, than I would otherwise be able to.