license: apache-2.0
tags:
- UNA
- simple-math
- juanako
base_model: abacusai/Smaug-34B-v0.1
datasets:
- fblgit/simple-math
- jondurbin/bagel-v0.3
model-index:
- name: UNA-SimpleSmaug-34b-v1beta
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: 74.57
name: normalized accuracy
source:
url: >-
https://huggingface.co/spaces/HuggingFaceH4/open_llm_leaderboard?query=fblgit/UNA-SimpleSmaug-34b-v1beta
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: 86.74
name: normalized accuracy
source:
url: >-
https://huggingface.co/spaces/HuggingFaceH4/open_llm_leaderboard?query=fblgit/UNA-SimpleSmaug-34b-v1beta
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: 76.68
name: accuracy
source:
url: >-
https://huggingface.co/spaces/HuggingFaceH4/open_llm_leaderboard?query=fblgit/UNA-SimpleSmaug-34b-v1beta
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: 70.17
source:
url: >-
https://huggingface.co/spaces/HuggingFaceH4/open_llm_leaderboard?query=fblgit/UNA-SimpleSmaug-34b-v1beta
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: 83.82
name: accuracy
source:
url: >-
https://huggingface.co/spaces/HuggingFaceH4/open_llm_leaderboard?query=fblgit/UNA-SimpleSmaug-34b-v1beta
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: 72.48
name: accuracy
source:
url: >-
https://huggingface.co/spaces/HuggingFaceH4/open_llm_leaderboard?query=fblgit/UNA-SimpleSmaug-34b-v1beta
name: Open LLM Leaderboard
UNA-SimpleSmaug-34b-v1beta
Scoring 04-February-2024 #1 34B model, outperforming its original base model Smaug-34B-v0.1 with 77.41 😎
Oh, btw.. this one went thru SFT so the abacus inside Smaug is back to normal.. so you can further train/dpo him .. RESET!
Applied UNA only on the Attention, not on the MLP's
- Is based on Smaug
- SimpleMath dataset
- It was trained on Axolotl
Experiment
The thing here is to understand whats the impact of SimpleMath applied at the attention layer during a SFT session and how it impacts on the neural network overall.
Results: Improving mathematican and reasoning capabilities without degrading and presserving previous training sessions.
And enjoy our ModelSimilarities tool detector https://github.com/fblgit/model-similarity where we confirmed numerically the bloodties of the model.
Evals
Pending, but so far this one
| Task |Version| Metric |Value |
|-------------|------:|--------|----------------:|
|arc_challenge| HF|acc_norm| 0.7457337883959 |
|gsm8k | HF|acc | 0.7247915087187 |
|mmlu | HF|acc | 0.7649553475572 |
|mmlu | HF|acc_norm| 0.7681713551647 |
|hellaswag | HF|acc_norm| 0.8673571001792 |
|truthfulqa | HF|mc2 | 0.7016557407771 |
|winogrande | HF|acc | 0.8382004735595 |
|------------------------------------------------|
Increasing GSM, MMLU, ARC, WINO.
Citations
To abacusai for making Smaug-34B, the Bagel, and all the magic behind the base model.
If you use the model, provide citation even for merges or anything.
Open LLM Leaderboard Evaluation Results
Detailed results can be found here
| Metric | Value |
|---|---|
| Avg. | 77.41 |
| AI2 Reasoning Challenge (25-Shot) | 74.57 |
| HellaSwag (10-Shot) | 86.74 |
| MMLU (5-Shot) | 76.68 |
| TruthfulQA (0-shot) | 70.17 |
| Winogrande (5-shot) | 83.82 |
| GSM8k (5-shot) | 72.48 |