thea-3b-25r / README.md
lunahr's picture
v2 is now available!
163a51e verified
|
raw
history blame
5.33 kB
---
language:
- en
license: llama3.2
tags:
- text-generation-inference
- transformers
- llama
- trl
- sft
- reasoning
- llama-3
base_model: chuanli11/Llama-3.2-3B-Instruct-uncensored
datasets:
- KingNish/reasoning-base-20k
- lunahr/thea-name-overrides
model-index:
- name: thea-3b-25r
results:
- task:
type: text-generation
name: Text Generation
dataset:
name: IFEval (0-Shot)
type: HuggingFaceH4/ifeval
args:
num_few_shot: 0
metrics:
- type: inst_level_strict_acc and prompt_level_strict_acc
value: 73.44
name: strict accuracy
source:
url: https://huggingface.co/spaces/open-llm-leaderboard/open_llm_leaderboard?query=piotr25691/thea-3b-25r
name: Open LLM Leaderboard
- task:
type: text-generation
name: Text Generation
dataset:
name: BBH (3-Shot)
type: BBH
args:
num_few_shot: 3
metrics:
- type: acc_norm
value: 22.55
name: normalized accuracy
source:
url: https://huggingface.co/spaces/open-llm-leaderboard/open_llm_leaderboard?query=piotr25691/thea-3b-25r
name: Open LLM Leaderboard
- task:
type: text-generation
name: Text Generation
dataset:
name: MATH Lvl 5 (4-Shot)
type: hendrycks/competition_math
args:
num_few_shot: 4
metrics:
- type: exact_match
value: 16.31
name: exact match
source:
url: https://huggingface.co/spaces/open-llm-leaderboard/open_llm_leaderboard?query=piotr25691/thea-3b-25r
name: Open LLM Leaderboard
- task:
type: text-generation
name: Text Generation
dataset:
name: GPQA (0-shot)
type: Idavidrein/gpqa
args:
num_few_shot: 0
metrics:
- type: acc_norm
value: 2.35
name: acc_norm
source:
url: https://huggingface.co/spaces/open-llm-leaderboard/open_llm_leaderboard?query=piotr25691/thea-3b-25r
name: Open LLM Leaderboard
- task:
type: text-generation
name: Text Generation
dataset:
name: MuSR (0-shot)
type: TAUR-Lab/MuSR
args:
num_few_shot: 0
metrics:
- type: acc_norm
value: 3.57
name: acc_norm
source:
url: https://huggingface.co/spaces/open-llm-leaderboard/open_llm_leaderboard?query=piotr25691/thea-3b-25r
name: Open LLM Leaderboard
- task:
type: text-generation
name: Text Generation
dataset:
name: MMLU-PRO (5-shot)
type: TIGER-Lab/MMLU-Pro
config: main
split: test
args:
num_few_shot: 5
metrics:
- type: acc
value: 24.25
name: accuracy
source:
url: https://huggingface.co/spaces/open-llm-leaderboard/open_llm_leaderboard?query=piotr25691/thea-3b-25r
name: Open LLM Leaderboard
---
# UPDATE AVAILABLE
A newer version is available. It is based on a better foundational model and it may provide higher benchmarking scores.
Check it out: https://huggingface.co/lunahr/thea-v2-3b-50r
# Model Description
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:
```py
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)
```
- **Trained by:** [Piotr Zalewski](https://huggingface.co/lunahr)
- **License:** llama3.2
- **Finetuned from model:** [chuanli11/Llama-3.2-3B-Instruct-uncensored](https://huggingface.co/chuanli11/Llama-3.2-3B-Instruct-uncensored)
- **Dataset used:** [KingNish/reasoning-base-20k](https://huggingface.co/datasets/KingNish/reasoning-base-20k)
This Llama model was trained faster than [Unsloth](https://github.com/unslothai/unsloth) using [custom training code](https://www.kaggle.com/code/piotr25691/distributed-llama-training-with-2xt4).
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.