c67c9f6d66
Low_mem doesn't work in exl2 and it was an experimental option to begin with. Keep the loading code commented out in case it gets fixed in the future. A better alternative is to use 8bit cache which works and helps save VRAM. Signed-off-by: kingbri <bdashore3@proton.me>
61 lines
2.1 KiB
YAML
61 lines
2.1 KiB
YAML
# Sample YAML file for configuration.
|
|
# Comment out values as needed. Every value has a default within the application.
|
|
|
|
# Unless specified in the comments, DO NOT put these options in quotes!
|
|
# You can use https://www.yamllint.com/ if you want to check your YAML formatting.
|
|
|
|
# Options for networking
|
|
network:
|
|
# The IP to host on (default: 127.0.0.1).
|
|
# Use 0.0.0.0 to expose on all network adapters
|
|
host: 127.0.0.1
|
|
|
|
# The port to host on (default: 5000)
|
|
port: 5000
|
|
|
|
# Options for model overrides and loading
|
|
model:
|
|
# Overrides the directory to look for models (default: models)
|
|
# Windows users, DO NOT put this path in quotes! This directory will be invalid otherwise.
|
|
model_dir: your model directory path
|
|
|
|
# An initial model to load. Make sure the model is located in the model directory!
|
|
# A model can be loaded later via the API.
|
|
model_name: A model name
|
|
|
|
# Sends dummy model names when the models endpoint is queried
|
|
# Enable this if the program is looking for a specific OAI model
|
|
use_dummy_models: False
|
|
|
|
# The below parameters apply only if model_name is set
|
|
|
|
# Maximum model context length (default: 4096)
|
|
max_seq_len: 4096
|
|
|
|
# Automatically allocate resources to GPUs (default: True)
|
|
gpu_split_auto: True
|
|
|
|
# An integer array of GBs of vram to split between GPUs (default: [])
|
|
gpu_split: [20.6, 24]
|
|
|
|
# Rope scaling parameters (default: 1.0)
|
|
rope_scale: 1.0
|
|
rope_alpha: 1.0
|
|
|
|
# Disable Flash-attention 2. Set to True for GPUs lower than Nvidia's 3000 series. (default: False)
|
|
no_flash_attention: False
|
|
|
|
# Enable 8 bit cache mode for VRAM savings (slight performance hit). Possible values FP16, FP8. (default: FP16)
|
|
cache_mode: FP16
|
|
|
|
# Options for draft models (speculative decoding). This will use more VRAM!
|
|
draft:
|
|
# Overrides the directory to look for draft (default: models)
|
|
draft_model_dir: Your draft model directory path
|
|
|
|
# An initial draft model to load. Make sure this model is located in the model directory!
|
|
# A draft model can be loaded later via the API.
|
|
draft_model_name: A model name
|
|
|
|
# Rope parameters for draft models (default: 1.0)
|
|
draft_rope_alpha: 1.0
|