Instructions to use IntervitensInc/Index-1.9b-glaive with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use IntervitensInc/Index-1.9b-glaive with Transformers:

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

pipe = pipeline("text-generation", model="IntervitensInc/Index-1.9b-glaive", trust_remote_code=True)
messages = [
    {"role": "user", "content": "Who are you?"},
]
pipe(messages)

# Load model directly
from transformers import AutoModelForCausalLM
model = AutoModelForCausalLM.from_pretrained("IntervitensInc/Index-1.9b-glaive", trust_remote_code=True, dtype="auto")

Notebooks
Google Colab
Kaggle
Local Apps Settings

vLLM

How to use IntervitensInc/Index-1.9b-glaive with vLLM:

Install from pip and serve model

# Install vLLM from pip:
pip install vllm
# Start the vLLM server:
vllm serve "IntervitensInc/Index-1.9b-glaive"
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:8000/v1/chat/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "IntervitensInc/Index-1.9b-glaive",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Use Docker

docker model run hf.co/IntervitensInc/Index-1.9b-glaive

SGLang

How to use IntervitensInc/Index-1.9b-glaive 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 "IntervitensInc/Index-1.9b-glaive" \
    --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": "IntervitensInc/Index-1.9b-glaive",
		"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 "IntervitensInc/Index-1.9b-glaive" \
        --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": "IntervitensInc/Index-1.9b-glaive",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Docker Model Runner
How to use IntervitensInc/Index-1.9b-glaive with Docker Model Runner:
```
docker model run hf.co/IntervitensInc/Index-1.9b-glaive
```

See axolotl config

axolotl version: 0.4.1

base_model: IndexTeam/Index-1.9B-Chat
model_type: IndexForCausalLM
tokenizer_type: AutoTokenizer

load_in_8bit: false
load_in_4bit: false
strict: false

datasets:
  - path: gardner/glaive-function-calling-v2-sharegpt
    type: sharegpt
    conversation: index_chat
dataset_prepared_path: last_run_prepared
val_set_size: 0.05
output_dir: ./outputs/out

sequence_len: 4096
sample_packing: true
pad_to_sequence_len: true
trust_remote_code: true

wandb_project: index-1.9b
wandb_entity:
wandb_watch:
wandb_name: index-1.9b
wandb_log_model:

gradient_accumulation_steps: 8
micro_batch_size: 1
num_epochs: 4
optimizer: paged_adamw_8bit
lr_scheduler: cosine
learning_rate: 5e-6

train_on_inputs: false
group_by_length: false
bf16: auto
fp16:
tf32: false

gradient_checkpointing: true
gradient_checkpointing_kwargs:
  use_reentrant: false
early_stopping_patience:
resume_from_checkpoint:
logging_steps: 1
xformers_attention:
flash_attention: true

warmup_steps: 100
evals_per_epoch: 2
eval_table_size:
saves_per_epoch: 2
debug:
deepspeed:
weight_decay: 0.0
fsdp:
fsdp_config:
special_tokens:
  pad_token: </s>

outputs/out

This model is a fine-tuned version of IndexTeam/Index-1.9B-Chat on the None dataset. It achieves the following results on the evaluation set:

Loss: 0.4745

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: 5e-06
train_batch_size: 1
eval_batch_size: 1
seed: 42
distributed_type: multi-GPU
num_devices: 4
gradient_accumulation_steps: 8
total_train_batch_size: 32
total_eval_batch_size: 4
optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
lr_scheduler_type: cosine
lr_scheduler_warmup_steps: 100
num_epochs: 4

Training results

Training Loss	Epoch	Step	Validation Loss
5.3244	0.0021	1	5.2299
0.6569	0.4992	237	0.6106
0.5053	0.9984	474	0.5233
0.472	1.4779	711	0.5007
0.4849	1.9771	948	0.4863
0.4611	2.4576	1185	0.4794
0.3926	2.9568	1422	0.4756
0.4171	3.4368	1659	0.4747
0.4246	3.9360	1896	0.4745

Framework versions

Transformers 4.42.3
Pytorch 2.1.2+cu118
Datasets 2.19.1
Tokenizers 0.19.1

Downloads last month: 5

Safetensors

Model size

2B params

Tensor type

BF16

Model tree for IntervitensInc/Index-1.9b-glaive

Base model

IndexTeam/Index-1.9B-Chat

Finetuned

(2)

this model