Skip to content

Commit

Permalink
[Model] Add SmolLM2, deprecate SmolLM1 (#623)
Browse files Browse the repository at this point in the history 
This PR adds the following SmolLM2 prebuilt models:
- SmolLM2-1.7B-Instruct-q4f16_1-MLC
- SmolLM2-1.7B-Instruct-q4f32_1-MLC
- SmolLM2-360M-Instruct-q0f16-MLC
- SmolLM2-360M-Instruct-q0f32-MLC
- SmolLM2-360M-Instruct-q4f16_1-MLC
- SmolLM2-360M-Instruct-q4f132_1-MLC
- SmolLM2-135M-Instruct-q0f16-MLC
- SmolLM2-135M-Instruct-q0f32-MLC

Also removes SmolLM 1 models
  • Loading branch information
@CharlieFRuan
CharlieFRuan authored Nov 4, 2024
1 parent 767e110 commit ddb1624
Showing 1 changed file with 41 additions and 95 deletions.
136 changes: 41 additions & 95 deletions src/config.ts
View file
Original file line number Diff line number Diff line change
Expand Up @@ -717,168 +717,114 @@ export const prebuiltAppConfig: AppConfig = {
sliding_window_size: -1,
},
},
// SmolLM
// SmolLM2
{
model: "https://huggingface.co/mlc-ai/SmolLM-1.7B-Instruct-q0f16-MLC",
model_id: "SmolLM-1.7B-Instruct-q0f16-MLC",
model: "https://huggingface.co/mlc-ai/SmolLM2-1.7B-Instruct-q4f16_1-MLC",
model_id: "SmolLM2-1.7B-Instruct-q4f16_1-MLC",
model_lib:
modelLibURLPrefix +
modelVersion +
"/SmolLM-1.7B-Instruct-q0f16-ctx2k_cs1k-webgpu.wasm",
vram_required_MB: 3736.19,
"/SmolLM2-1.7B-Instruct-q4f16_1-ctx4k_cs1k-webgpu.wasm",
vram_required_MB: 1774.19,
low_resource_required: true,
required_features: ["shader-f16"],
overrides: {
context_window_size: 2048,
},
},
{
model: "https://huggingface.co/mlc-ai/SmolLM-1.7B-Instruct-q0f32-MLC",
model_id: "SmolLM-1.7B-Instruct-q0f32-MLC",
model_lib:
modelLibURLPrefix +
modelVersion +
"/SmolLM-1.7B-Instruct-q0f32-ctx2k_cs1k-webgpu.wasm",
vram_required_MB: 7432.38,
low_resource_required: false,
overrides: {
context_window_size: 2048,
},
},
{
model: "https://huggingface.co/mlc-ai/SmolLM-1.7B-Instruct-q4f16_1-MLC",
model_id: "SmolLM-1.7B-Instruct-q4f16_1-MLC",
model_lib:
modelLibURLPrefix +
modelVersion +
"/SmolLM-1.7B-Instruct-q4f16_1-ctx2k_cs1k-webgpu.wasm",
vram_required_MB: 1390.19,
low_resource_required: true,
required_features: ["shader-f16"],
overrides: {
context_window_size: 2048,
context_window_size: 4096,
},
},
{
model: "https://huggingface.co/mlc-ai/SmolLM-1.7B-Instruct-q4f32_1-MLC",
model_id: "SmolLM-1.7B-Instruct-q4f32_1-MLC",
model: "https://huggingface.co/mlc-ai/SmolLM2-1.7B-Instruct-q4f32_1-MLC",
model_id: "SmolLM2-1.7B-Instruct-q4f32_1-MLC",
model_lib:
modelLibURLPrefix +
modelVersion +
"/SmolLM-1.7B-Instruct-q4f32_1-ctx2k_cs1k-webgpu.wasm",
vram_required_MB: 1924.38,
"/SmolLM2-1.7B-Instruct-q4f32_1-ctx4k_cs1k-webgpu.wasm",
vram_required_MB: 2692.38,
low_resource_required: true,
overrides: {
context_window_size: 2048,
context_window_size: 4096,
},
},

{
model: "https://huggingface.co/mlc-ai/SmolLM-360M-Instruct-q0f16-MLC",
model_id: "SmolLM-360M-Instruct-q0f16-MLC",
model_lib:
modelLibURLPrefix +
modelVersion +
"/SmolLM-360M-Instruct-q0f16-ctx2k_cs1k-webgpu.wasm",
vram_required_MB: 791.99,
low_resource_required: true,
required_features: ["shader-f16"],
overrides: {
context_window_size: 2048,
},
},
{
model: "https://huggingface.co/mlc-ai/SmolLM-360M-Instruct-q0f32-MLC",
model_id: "SmolLM-360M-Instruct-q0f32-MLC",
model: "https://huggingface.co/mlc-ai/SmolLM2-360M-Instruct-q0f16-MLC",
model_id: "SmolLM2-360M-Instruct-q0f16-MLC",
model_lib:
modelLibURLPrefix +
modelVersion +
"/SmolLM-360M-Instruct-q0f32-ctx2k_cs1k-webgpu.wasm",
vram_required_MB: 1583.99,
low_resource_required: true,
overrides: {
context_window_size: 2048,
},
},
{
model: "https://huggingface.co/mlc-ai/SmolLM-360M-Instruct-q4f16_1-MLC",
model_id: "SmolLM-360M-Instruct-q4f16_1-MLC",
model_lib:
modelLibURLPrefix +
modelVersion +
"/SmolLM-360M-Instruct-q4f16_1-ctx2k_cs1k-webgpu.wasm",
vram_required_MB: 296.06,
"/SmolLM2-360M-Instruct-q0f16-ctx4k_cs1k-webgpu.wasm",
vram_required_MB: 871.99,
low_resource_required: true,
required_features: ["shader-f16"],
overrides: {
context_window_size: 2048,
context_window_size: 4096,
},
},
{
model: "https://huggingface.co/mlc-ai/SmolLM-360M-Instruct-q4f32_1-MLC",
model_id: "SmolLM-360M-Instruct-q4f32_1-MLC",
model: "https://huggingface.co/mlc-ai/SmolLM2-360M-Instruct-q0f32-MLC",
model_id: "SmolLM2-360M-Instruct-q0f32-MLC",
model_lib:
modelLibURLPrefix +
modelVersion +
"/SmolLM-360M-Instruct-q4f32_1-ctx2k_cs1k-webgpu.wasm",
vram_required_MB: 419.61,
"/SmolLM2-360M-Instruct-q0f32-ctx4k_cs1k-webgpu.wasm",
vram_required_MB: 1743.99,
low_resource_required: true,
overrides: {
context_window_size: 2048,
context_window_size: 4096,
},
},
{
model: "https://huggingface.co/mlc-ai/SmolLM-135M-Instruct-q0f16-MLC",
model_id: "SmolLM-135M-Instruct-q0f16-MLC",
model: "https://huggingface.co/mlc-ai/SmolLM2-360M-Instruct-q4f16_1-MLC",
model_id: "SmolLM2-360M-Instruct-q4f16_1-MLC",
model_lib:
modelLibURLPrefix +
modelVersion +
"/SmolLM-135M-Instruct-q0f16-ctx2k_cs1k-webgpu.wasm",
vram_required_MB: 314.69,
"/SmolLM2-360M-Instruct-q4f16_1-ctx4k_cs1k-webgpu.wasm",
vram_required_MB: 376.06,
low_resource_required: true,
required_features: ["shader-f16"],
overrides: {
context_window_size: 2048,
context_window_size: 4096,
},
},
{
model: "https://huggingface.co/mlc-ai/SmolLM-135M-Instruct-q0f32-MLC",
model_id: "SmolLM-135M-Instruct-q0f32-MLC",
model: "https://huggingface.co/mlc-ai/SmolLM2-360M-Instruct-q4f32_1-MLC",
model_id: "SmolLM2-360M-Instruct-q4f32_1-MLC",
model_lib:
modelLibURLPrefix +
modelVersion +
"/SmolLM-135M-Instruct-q0f32-ctx2k_cs1k-webgpu.wasm",
vram_required_MB: 629.38,
"/SmolLM2-360M-Instruct-q4f32_1-ctx4k_cs1k-webgpu.wasm",
vram_required_MB: 579.61,
low_resource_required: true,
overrides: {
context_window_size: 2048,
context_window_size: 4096,
},
},
{
model: "https://huggingface.co/mlc-ai/SmolLM-135M-Instruct-q4f16_1-MLC",
model_id: "SmolLM-135M-Instruct-q4f16_1-MLC",
model: "https://huggingface.co/mlc-ai/SmolLM2-135M-Instruct-q0f16-MLC",
model_id: "SmolLM2-135M-Instruct-q0f16-MLC",
model_lib:
modelLibURLPrefix +
modelVersion +
"/SmolLM-135M-Instruct-q4f16_1-ctx2k_cs1k-webgpu.wasm",
vram_required_MB: 130.33,
"/SmolLM2-135M-Instruct-q0f16-ctx4k_cs1k-webgpu.wasm",
vram_required_MB: 359.69,
low_resource_required: true,
required_features: ["shader-f16"],
overrides: {
context_window_size: 2048,
context_window_size: 4096,
},
},
{
model: "https://huggingface.co/mlc-ai/SmolLM-135M-Instruct-q4f32_1-MLC",
model_id: "SmolLM-135M-Instruct-q4f32_1-MLC",
model: "https://huggingface.co/mlc-ai/SmolLM2-135M-Instruct-q0f32-MLC",
model_id: "SmolLM2-135M-Instruct-q0f32-MLC",
model_lib:
modelLibURLPrefix +
modelVersion +
"/SmolLM-135M-Instruct-q4f32_1-ctx2k_cs1k-webgpu.wasm",
vram_required_MB: 196.54,
"/SmolLM2-135M-Instruct-q0f32-ctx4k_cs1k-webgpu.wasm",
vram_required_MB: 719.38,
low_resource_required: true,
overrides: {
context_window_size: 2048,
context_window_size: 4096,
},
},
// Gemma2
Expand Down

0 comments on commit ddb1624

Please sign in to comment.