Instructions to use tsavage68/chat_550_STEPS_01beta_1e6_rate_CDPOSFT with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use tsavage68/chat_550_STEPS_01beta_1e6_rate_CDPOSFT with Transformers:

# Use a pipeline as a high-level helper
from transformers import pipeline

pipe = pipeline("text-generation", model="tsavage68/chat_550_STEPS_01beta_1e6_rate_CDPOSFT")
messages = [
    {"role": "user", "content": "Who are you?"},
]
pipe(messages)

# Load model directly
from transformers import AutoTokenizer, AutoModelForCausalLM

tokenizer = AutoTokenizer.from_pretrained("tsavage68/chat_550_STEPS_01beta_1e6_rate_CDPOSFT")
model = AutoModelForCausalLM.from_pretrained("tsavage68/chat_550_STEPS_01beta_1e6_rate_CDPOSFT", device_map="auto")
messages = [
    {"role": "user", "content": "Who are you?"},
]
inputs = tokenizer.apply_chat_template(
	messages,
	add_generation_prompt=True,
	tokenize=True,
	return_dict=True,
	return_tensors="pt",
).to(model.device)

outputs = model.generate(**inputs, max_new_tokens=40)
print(tokenizer.decode(outputs[0][inputs["input_ids"].shape[-1]:]))

Notebooks
Google Colab
Kaggle
Local Apps Settings

vLLM

How to use tsavage68/chat_550_STEPS_01beta_1e6_rate_CDPOSFT with vLLM:

Install from pip and serve model

# Install vLLM from pip:
pip install vllm
# Start the vLLM server:
vllm serve "tsavage68/chat_550_STEPS_01beta_1e6_rate_CDPOSFT"
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:8000/v1/chat/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "tsavage68/chat_550_STEPS_01beta_1e6_rate_CDPOSFT",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Use Docker

docker model run hf.co/tsavage68/chat_550_STEPS_01beta_1e6_rate_CDPOSFT

SGLang

How to use tsavage68/chat_550_STEPS_01beta_1e6_rate_CDPOSFT with SGLang:

Install from pip and serve model

# Install SGLang from pip:
pip install sglang
# Start the SGLang server:
python3 -m sglang.launch_server \
    --model-path "tsavage68/chat_550_STEPS_01beta_1e6_rate_CDPOSFT" \
    --host 0.0.0.0 \
    --port 30000
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:30000/v1/chat/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "tsavage68/chat_550_STEPS_01beta_1e6_rate_CDPOSFT",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Use Docker images

docker run --gpus all \
    --shm-size 32g \
    -p 30000:30000 \
    -v ~/.cache/huggingface:/root/.cache/huggingface \
    --env "HF_TOKEN=<secret>" \
    --ipc=host \
    lmsysorg/sglang:latest \
    python3 -m sglang.launch_server \
        --model-path "tsavage68/chat_550_STEPS_01beta_1e6_rate_CDPOSFT" \
        --host 0.0.0.0 \
        --port 30000
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:30000/v1/chat/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "tsavage68/chat_550_STEPS_01beta_1e6_rate_CDPOSFT",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Docker Model Runner
How to use tsavage68/chat_550_STEPS_01beta_1e6_rate_CDPOSFT with Docker Model Runner:
```
docker model run hf.co/tsavage68/chat_550_STEPS_01beta_1e6_rate_CDPOSFT
```

chat_550_STEPS_01beta_1e6_rate_CDPOSFT

This model is a fine-tuned version of tsavage68/chat_600STEPS_1e8rate_SFT on an unknown dataset. It achieves the following results on the evaluation set:

Loss: 0.6716
Rewards/chosen: -0.1192
Rewards/rejected: -0.1802
Rewards/accuracies: 0.5253
Rewards/margins: 0.0610
Logps/rejected: -20.6044
Logps/chosen: -17.9469
Logits/rejected: -0.6222
Logits/chosen: -0.6220

Model description

More information needed

Intended uses & limitations

More information needed

Training and evaluation data

More information needed

Training procedure

Training hyperparameters

The following hyperparameters were used during training:

learning_rate: 1e-06
train_batch_size: 4
eval_batch_size: 1
seed: 42
gradient_accumulation_steps: 2
total_train_batch_size: 8
optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
lr_scheduler_type: cosine
lr_scheduler_warmup_steps: 100
training_steps: 550

Training results

Training Loss	Epoch	Step	Validation Loss	Rewards/chosen	Rewards/rejected	Rewards/accuracies	Rewards/margins	Logps/rejected	Logps/chosen	Logits/rejected	Logits/chosen
0.6925	0.0977	50	0.6917	0.0117	0.0085	0.4659	0.0031	-18.7166	-16.6380	-0.6015	-0.6013
0.6776	0.1953	100	0.6812	-0.0371	-0.0646	0.5253	0.0275	-19.4479	-17.1259	-0.6242	-0.6241
0.6927	0.2930	150	0.6819	-0.0802	-0.1112	0.5011	0.0310	-19.9140	-17.5569	-0.6222	-0.6220
0.6928	0.3906	200	0.6776	-0.1032	-0.1444	0.5033	0.0412	-20.2463	-17.7865	-0.6050	-0.6048
0.6937	0.4883	250	0.6762	-0.0643	-0.1121	0.5121	0.0478	-19.9228	-17.3977	-0.6013	-0.6011
0.6758	0.5859	300	0.6717	-0.1055	-0.1663	0.5231	0.0608	-20.4645	-17.8094	-0.6301	-0.6299
0.6696	0.6836	350	0.6724	-0.1144	-0.1731	0.5275	0.0587	-20.5330	-17.8991	-0.6162	-0.6160
0.6587	0.7812	400	0.6711	-0.1221	-0.1842	0.5297	0.0621	-20.6441	-17.9756	-0.6249	-0.6247
0.6755	0.8789	450	0.6713	-0.1178	-0.1794	0.5341	0.0616	-20.5960	-17.9326	-0.6214	-0.6212
0.6637	0.9766	500	0.6712	-0.1188	-0.1808	0.5253	0.0620	-20.6100	-17.9427	-0.6222	-0.6220
0.5575	1.0742	550	0.6716	-0.1192	-0.1802	0.5253	0.0610	-20.6044	-17.9469	-0.6222	-0.6220

Framework versions

Transformers 4.40.1
Pytorch 2.0.0+cu117
Datasets 2.19.0
Tokenizers 0.19.1

Downloads last month: 6

Safetensors

Model size

7B params

Tensor type

F16

Model tree for tsavage68/chat_550_STEPS_01beta_1e6_rate_CDPOSFT

Base model

meta-llama/Llama-2-7b-chat-hf

Finetuned

tsavage68/chat_600STEPS_1e8rate_SFT

Finetuned

(23)

this model