language:
- en
license: apache-2.0
datasets:
- cerebras/SlimPajama-627B
- bigcode/starcoderdata
- HuggingFaceH4/ultrachat_200k
- HuggingFaceH4/ultrafeedback_binarized
model-index:
- name: TinyLlama-1.1B-Remix-V.2
results:
- task:
type: text-generation
name: Text Generation
dataset:
name: AI2 Reasoning Challenge (25-Shot)
type: ai2_arc
config: ARC-Challenge
split: test
args:
num_few_shot: 25
metrics:
- type: acc_norm
value: 33.19
name: normalized accuracy
source:
url: >-
https://huggingface.co/spaces/HuggingFaceH4/open_llm_leaderboard?query=Deathsquad10/TinyLlama-1.1B-Remix-V.2
name: Open LLM Leaderboard
- task:
type: text-generation
name: Text Generation
dataset:
name: HellaSwag (10-Shot)
type: hellaswag
split: validation
args:
num_few_shot: 10
metrics:
- type: acc_norm
value: 56.62
name: normalized accuracy
source:
url: >-
https://huggingface.co/spaces/HuggingFaceH4/open_llm_leaderboard?query=Deathsquad10/TinyLlama-1.1B-Remix-V.2
name: Open LLM Leaderboard
- task:
type: text-generation
name: Text Generation
dataset:
name: MMLU (5-Shot)
type: cais/mmlu
config: all
split: test
args:
num_few_shot: 5
metrics:
- type: acc
value: 25.99
name: accuracy
source:
url: >-
https://huggingface.co/spaces/HuggingFaceH4/open_llm_leaderboard?query=Deathsquad10/TinyLlama-1.1B-Remix-V.2
name: Open LLM Leaderboard
- task:
type: text-generation
name: Text Generation
dataset:
name: TruthfulQA (0-shot)
type: truthful_qa
config: multiple_choice
split: validation
args:
num_few_shot: 0
metrics:
- type: mc2
value: 34.64
source:
url: >-
https://huggingface.co/spaces/HuggingFaceH4/open_llm_leaderboard?query=Deathsquad10/TinyLlama-1.1B-Remix-V.2
name: Open LLM Leaderboard
- task:
type: text-generation
name: Text Generation
dataset:
name: Winogrande (5-shot)
type: winogrande
config: winogrande_xl
split: validation
args:
num_few_shot: 5
metrics:
- type: acc
value: 58.09
name: accuracy
source:
url: >-
https://huggingface.co/spaces/HuggingFaceH4/open_llm_leaderboard?query=Deathsquad10/TinyLlama-1.1B-Remix-V.2
name: Open LLM Leaderboard
- task:
type: text-generation
name: Text Generation
dataset:
name: GSM8k (5-shot)
type: gsm8k
config: main
split: test
args:
num_few_shot: 5
metrics:
- type: acc
value: 0.91
name: accuracy
source:
url: >-
https://huggingface.co/spaces/HuggingFaceH4/open_llm_leaderboard?query=Deathsquad10/TinyLlama-1.1B-Remix-V.2
name: Open LLM Leaderboard
| Tasks | Version | Filter | n-shot | Metric | Value | Stderr | |
|---|---|---|---|---|---|---|---|
| arc_challenge | Yaml | none | 0 | acc | 0.2790 | ± | 0.0131 |
| none | 0 | acc_norm | 0.3234 | ± | 0.0137 | ||
| arc_easy | Yaml | none | 0 | acc | 0.6006 | ± | 0.0101 |
| none | 0 | acc_norm | 0.5770 | ± | 0.0101 | ||
| boolq | Yaml | none | 0 | acc | 0.6373 | ± | 0.0084 |
| hellaswag | Yaml | none | 0 | acc | 0.4521 | ± | 0.0050 |
| none | 0 | acc_norm | 0.5822 | ± | 0.0049 | ||
| openbookqa | Yaml | none | 0 | acc | 0.2220 | ± | 0.0186 |
| none | 0 | acc_norm | 0.3740 | ± | 0.0217 | ||
| piqa | Yaml | none | 0 | acc | 0.7269 | ± | 0.0104 |
| none | 0 | acc_norm | 0.7296 | ± | 0.0104 | ||
| winogrande | Yaml | none | 0 | acc | 0.5754 | ± | 0.0139 |
https://github.com/jzhang38/TinyLlama
The TinyLlama project aims to pretrain a 1.1B Llama model on 3 trillion tokens. With some proper optimization, we can achieve this within a span of "just" 90 days using 16 A100-40G GPUs 🚀🚀. The training has started on 2023-09-01.
We adopted exactly the same architecture and tokenizer as Llama 2. This means TinyLlama can be plugged and played in many open-source projects built upon Llama. Besides, TinyLlama is compact with only 1.1B parameters. This compactness allows it to cater to a multitude of applications demanding a restricted computation and memory footprint.
This Model
This is the chat model finetuned on top of TinyLlama/TinyLlama-1.1B-intermediate-step-1431k-3T. We follow HF's Zephyr's training recipe. The model was " initially fine-tuned on a variant of the UltraChat dataset, which contains a diverse range of synthetic dialogues generated by ChatGPT.
We then further aligned the model with 🤗 TRL's DPOTrainer on the openbmb/UltraFeedback dataset, which contain 64k prompts and model completions that are ranked by GPT-4."
How to use
You will need the transformers>=4.34 Do check the TinyLlama github page for more information.
# Install transformers from source - only needed for versions <= v4.34
# pip install git+https://github.com/huggingface/transformers.git
# pip install accelerate
import torch
from transformers import pipeline
pipe = pipeline("text-generation", model="TinyLlama/TinyLlama-1.1B-Chat-v1.0", torch_dtype=torch.bfloat16, device_map="auto")
# We use the tokenizer's chat template to format each message - see https://huggingface.co/docs/transformers/main/en/chat_templating
messages = [
{
"role": "system",
"content": "You are a friendly chatbot who always responds in the style of a pirate",
},
{"role": "user", "content": "How many helicopters can a human eat in one sitting?"},
]
prompt = pipe.tokenizer.apply_chat_template(messages, tokenize=False, add_generation_prompt=True)
outputs = pipe(prompt, max_new_tokens=256, do_sample=True, temperature=0.7, top_k=50, top_p=0.95)
print(outputs[0]["generated_text"])
# <|system|>
# You are a friendly chatbot who always responds in the style of a pirate.</s>
# <|user|>
# How many helicopters can a human eat in one sitting?</s>
# <|assistant|>
# ...
Open LLM Leaderboard Evaluation Results
Detailed results can be found here
| Metric | Value |
|---|---|
| Avg. | 34.91 |
| AI2 Reasoning Challenge (25-Shot) | 33.19 |
| HellaSwag (10-Shot) | 56.62 |
| MMLU (5-Shot) | 25.99 |
| TruthfulQA (0-shot) | 34.64 |
| Winogrande (5-shot) | 58.09 |
| GSM8k (5-shot) | 0.91 |