Downloads · 30 days
81
0% of all-time downloads
speechless-tora-code-7b-v1.0
uukuguy/speechless-tora-code-7b-v1.0
speechless-tora-code-7b-v1.0 is a text generation model from uukuguy. Use it when you need the model to write or continue text. It is set up for transformers. The card lists the license as llama2.
<p<h1 speechless-tora-code-7b-v1.0 </h1</p
All-time downloads
53.4K
Public
Repo size
28.6 GB
Likes
2
Public
Repo makeup
Click a slice to open those files.
.bin15.4 GB · 98%
At a glance
- Task
- Text Generation
- Library
- transformers
- License
- llama2
- Model type
- llama
- Access
- Public
- Created
- Oct 10, 2023
- Updated
- Dec 30, 2023
- SHA
- 0a1f7a53
Try a prompt
Model card
From the Hugging Face model README
<p><h1> speechless-tora-code-7b-v1.0 </h1></p>
- AWQ model(s) for GPU inference.
- GPTQ models for GPU inference, with multiple quantisation parameter options.
- 2, 3, 4, 5, 6 and 8-bit GGUF models for CPU+GPU inference
Code: https://github.com/uukuguy/speechless
Use the following dataset to fine-tune llm_agents/tora-code-7b-v1.0 in order to improve the model's reasoning and planning abilities.
Total 201,981 samples.
- jondurbin/airoboros-2.2: Filter categories related to coding, reasoning and planning. 23,462 samples.
- Open-Orca/OpenOrca: Filter the 'cot' category in 1M GPT4 dataset. 74,440 samples.
- garage-bAInd/Open-Platypus: 100%, 24,926 samples.
- WizardLM/WizardLM_evol_instruct_V2_196k: Coding coversation part. 30,185 samples
- TokenBender/python_eval_instruct_51k: “python” in output .40,309 samples
- Spider: 8,659 samples
How to Prompt the Model
This model accepts the Alpaca instruction format.
For example:
You are an intelligent programming assistant.
### Instruction:
Implement a linked list in C++
### Response:
HumanEval
| Metric | Value |
|---|---|
| humaneval-python | 51.829 |
CodeLlama-34B-Python: 53.29
CodeLlama-34B-Instruct: 50.79
CodeLlama-13B-Instruct: 50.6
CodeLlama-34B: 45.11
CodeLlama-13B-Python: 42.89
CodeLlama-13B: 35.07
LM-Evaluation-Harness
| Metric | Value |
|---|---|
| ARC | 42.66 |
| HellaSwag | 65.16 |
| MMLU | 38.56 |
| TruthfulQA | 42.06 |
| Average | 47.11 |
Parameters
| lr | 2e-4 |
| lr_scheduler_type | cosine |
| weight_decay | 0.0 |
| optim | paged_adamw_8bit |
| flash_attention | True |
| rerope | False |
| max_new_tokens | 4096 |
| num_train_epochs | 2 |
| bits | 4 |
| lora_r | 64 |
| lora_alpha | 16 |
| lora_dropout | 0.05 |
| double_quant | True |
| quant_type | nf4 |
| dataset_format | airoboros |
| mini_batch_size | 2 |
| grandient_accumulation_steps | 32 |
| bf16 | True |
A800-80G x 2
| epoch | 2.0 |
| etrain_loss | 0.5891 |
| etrain_runtime | 19:24:49.43 |
| etrain_samples_per_second | 5.664 |
| etrain_steps_per_second | 0.044 |
| eeval_loss | 0.5872 |
| eeval_runtime | 0:00:15.59 |
| eeval_samples_per_second | 12.822 |
| eeval_steps_per_second | 6.411 |
Open LLM Leaderboard Evaluation Results
Detailed results can be found here
| Metric | Value |
|---|---|
| Avg. | 40.1 |
| ARC (25-shot) | 42.66 |
| HellaSwag (10-shot) | 65.16 |
| MMLU (5-shot) | 38.56 |
| TruthfulQA (0-shot) | 42.06 |
| Winogrande (5-shot) | 62.9 |
| GSM8K (5-shot) | 0.91 |
| DROP (3-shot) | 28.48 |
- Task
- Text Generation
- Library
- transformers
- Type
- llama
- License
- llama2
- Languages
- en
- Created
- Oct 10, 2023
- Updated
- Dec 30, 2023