Instructions to use dphn/DolphinHermes-120b with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use dphn/DolphinHermes-120b with Transformers:

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

pipe = pipeline("text-generation", model="dphn/DolphinHermes-120b")
messages = [
    {"role": "user", "content": "Who are you?"},
]
pipe(messages)

# Load model directly
from transformers import AutoTokenizer, AutoModelForCausalLM

tokenizer = AutoTokenizer.from_pretrained("dphn/DolphinHermes-120b")
model = AutoModelForCausalLM.from_pretrained("dphn/DolphinHermes-120b")
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

vLLM

How to use dphn/DolphinHermes-120b with vLLM:

Install from pip and serve model

# Install vLLM from pip:
pip install vllm
# Start the vLLM server:
vllm serve "dphn/DolphinHermes-120b"
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:8000/v1/chat/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "dphn/DolphinHermes-120b",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Use Docker

docker model run hf.co/dphn/DolphinHermes-120b

SGLang

How to use dphn/DolphinHermes-120b 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 "dphn/DolphinHermes-120b" \
    --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": "dphn/DolphinHermes-120b",
		"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 "dphn/DolphinHermes-120b" \
        --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": "dphn/DolphinHermes-120b",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Docker Model Runner
How to use dphn/DolphinHermes-120b with Docker Model Runner:
```
docker model run hf.co/dphn/DolphinHermes-120b
```
Browse Quantizations to use this model in llama.cpp, Ollama, LM Studio, or any compatible app.

DolphinHermes-120b

Cheers @teknium

This is a merge of pre-trained language models created using mergekit.

Discord: https://discord.gg/cognitivecomputations

Merge Details

Merge Method

This model was merged using the linear merge method.

Models Merged

The following models were included in the merge:

Configuration

The following YAML configuration was used to produce this model:

merge_method: linear # use linear so we can include multiple models, albeit at a zero weight
parameters:
  weight: 1.0 # weight everything as 1 unless specified otherwise - linear with one model weighted at 1 is a no-op like passthrough
slices:
  - sources:
      - model: cognitivecomputations/dolphin-2.2-70b # embed_tokens comes along with the ride with whatever is the first layer
        layer_range: [0, 1]
      - model: NousResearch/Nous-Hermes-2-Llama-2-70B # add dummy second model with 0 weight so tokenizer-based merge routine is invoked for embed_tokens
        layer_range: [0, 1]
        parameters:
          weight: 0
  - sources:
      - model: cognitivecomputations/dolphin-2.2-70b
        layer_range: [1, 20]
  - sources:
      - model: NousResearch/Nous-Hermes-2-Llama-2-70B
        layer_range: [10, 30]
  - sources:
      - model: cognitivecomputations/dolphin-2.2-70b
        layer_range: [20, 40]
  - sources:
      - model: NousResearch/Nous-Hermes-2-Llama-2-70B
        layer_range: [30, 50]
  - sources:
      - model: cognitivecomputations/dolphin-2.2-70b
        layer_range: [40, 60]
  - sources:
      - model: NousResearch/Nous-Hermes-2-Llama-2-70B
        layer_range: [50, 70]
  - sources:
      - model: cognitivecomputations/dolphin-2.2-70b
        layer_range: [60, 79]
  - sources: # same as above, but for lm_head with the last layer
      - model: cognitivecomputations/dolphin-2.2-70b
        layer_range: [79, 80]
      - model: NousResearch/Nous-Hermes-2-Llama-2-70B
        layer_range: [79, 80]
        parameters:
          weight: 0
dtype: float16
tokenizer_source: model:cognitivecomputations/dolphin-2.2-70b # keep exact tokenizer used by dolphin - or you could use `union` if you add all of the input models to the first/last slice, but they would need to be non-zero weight or you'll get NaNs in your embeddings

Downloads last month: 11

Safetensors

Model size

120B params

Tensor type

F16

Model tree for dphn/DolphinHermes-120b

NousResearch/Nous-Hermes-2-Llama-2-70B

dphn/dolphin-2.2-70b

Merge model

this model

Quantizations

2 models

Paper for dphn/DolphinHermes-120b

Model soups: averaging weights of multiple fine-tuned models improves accuracy without increasing inference time

Paper • 2203.05482 • Published Mar 10, 2022 • 8