web-ai-toolkit
Version:
AI powered features on the web made easy
1,350 lines • 5.26 MB
JavaScript
const js = "MLC_DUMMY_REQUIRE_VAR", xs = "MLC_DUMMY_REQUIRE_VAR";
var cD = typeof globalThis < "u" ? globalThis : typeof window < "u" ? window : typeof global < "u" ? global : typeof self < "u" ? self : {};
function Ts(B) {
if (B.__esModule) return B;
var A = Object.defineProperty({}, "__esModule", { value: !0 });
return Object.keys(B).forEach(function(g) {
var Q = Object.getOwnPropertyDescriptor(B, g);
Object.defineProperty(A, g, Q.get ? Q : {
enumerable: !0,
get: function() {
return B[g];
}
});
}), A;
}
var Ui = { exports: {} };
(function(B) {
(function(A, g) {
B.exports ? B.exports = g() : A.log = g();
})(cD, function() {
var A = function() {
}, g = "undefined", Q = typeof window !== g && typeof window.navigator !== g && /Trident\/|MSIE /.test(window.navigator.userAgent), I = [
"trace",
"debug",
"info",
"warn",
"error"
], C = {}, D = null;
function a(u, q) {
var H = u[q];
if (typeof H.bind == "function")
return H.bind(u);
try {
return Function.prototype.bind.call(H, u);
} catch {
return function() {
return Function.prototype.apply.apply(H, [u, arguments]);
};
}
}
function e() {
console.log && (console.log.apply ? console.log.apply(console, arguments) : Function.prototype.apply.apply(console.log, [console, arguments])), console.trace && console.trace();
}
function Y(u) {
return u === "debug" && (u = "log"), typeof console === g ? !1 : u === "trace" && Q ? e : console[u] !== void 0 ? a(console, u) : console.log !== void 0 ? a(console, "log") : A;
}
function U() {
for (var u = this.getLevel(), q = 0; q < I.length; q++) {
var H = I[q];
this[H] = q < u ? A : this.methodFactory(H, u, this.name);
}
if (this.log = this.debug, typeof console === g && u < this.levels.SILENT)
return "No console available for logging";
}
function p(u) {
return function() {
typeof console !== g && (U.call(this), this[u].apply(this, arguments));
};
}
function r(u, q, H) {
return Y(u) || p.apply(this, arguments);
}
function d(u, q) {
var H = this, z, W, X, IA = "loglevel";
typeof u == "string" ? IA += ":" + u : typeof u == "symbol" && (IA = void 0);
function tA(XA) {
var Rg = (I[XA] || "silent").toUpperCase();
if (!(typeof window === g || !IA)) {
try {
window.localStorage[IA] = Rg;
return;
} catch {
}
try {
window.document.cookie = encodeURIComponent(IA) + "=" + Rg + ";";
} catch {
}
}
}
function MA() {
var XA;
if (!(typeof window === g || !IA)) {
try {
XA = window.localStorage[IA];
} catch {
}
if (typeof XA === g)
try {
var Rg = window.document.cookie, _g = encodeURIComponent(IA), KA = Rg.indexOf(_g + "=");
KA !== -1 && (XA = /^([^;]+)/.exec(
Rg.slice(KA + _g.length + 1)
)[1]);
} catch {
}
return H.levels[XA] === void 0 && (XA = void 0), XA;
}
}
function aA() {
if (!(typeof window === g || !IA)) {
try {
window.localStorage.removeItem(IA);
} catch {
}
try {
window.document.cookie = encodeURIComponent(IA) + "=; expires=Thu, 01 Jan 1970 00:00:00 UTC";
} catch {
}
}
}
function fA(XA) {
var Rg = XA;
if (typeof Rg == "string" && H.levels[Rg.toUpperCase()] !== void 0 && (Rg = H.levels[Rg.toUpperCase()]), typeof Rg == "number" && Rg >= 0 && Rg <= H.levels.SILENT)
return Rg;
throw new TypeError("log.setLevel() called with invalid level: " + XA);
}
H.name = u, H.levels = {
TRACE: 0,
DEBUG: 1,
INFO: 2,
WARN: 3,
ERROR: 4,
SILENT: 5
}, H.methodFactory = q || r, H.getLevel = function() {
return X ?? W ?? z;
}, H.setLevel = function(XA, Rg) {
return X = fA(XA), Rg !== !1 && tA(X), U.call(H);
}, H.setDefaultLevel = function(XA) {
W = fA(XA), MA() || H.setLevel(XA, !1);
}, H.resetLevel = function() {
X = null, aA(), U.call(H);
}, H.enableAll = function(XA) {
H.setLevel(H.levels.TRACE, XA);
}, H.disableAll = function(XA) {
H.setLevel(H.levels.SILENT, XA);
}, H.rebuild = function() {
if (D !== H && (z = fA(D.getLevel())), U.call(H), D === H)
for (var XA in C)
C[XA].rebuild();
}, z = fA(
D ? D.getLevel() : "WARN"
);
var qg = MA();
qg != null && (X = fA(qg)), U.call(H);
}
D = new d(), D.getLogger = function(q) {
if (typeof q != "symbol" && typeof q != "string" || q === "")
throw new TypeError("You must supply a name when creating a logger.");
var H = C[q];
return H || (H = C[q] = new d(
q,
D.methodFactory
)), H;
};
var f = typeof window !== g ? window.log : void 0;
return D.noConflict = function() {
return typeof window !== g && window.log === D && (window.log = f), D;
}, D.getLoggers = function() {
return C;
}, D.default = D, D;
});
})(Ui);
var Og = Ui.exports;
class Ps extends Error {
constructor(A) {
super(`Cannot find model record in appConfig for ${A}. Please check if the model ID is correct and included in the model_list configuration.`), this.name = "ModelNotFoundError";
}
}
class SE extends Error {
constructor(A) {
super(A), this.name = "ConfigValueError";
}
}
class wI extends SE {
constructor(A, g) {
super(`Make sure \`${A}\` > ${g}.`), this.name = "MinValueError";
}
}
class nI extends SE {
constructor(A, g, Q, I) {
super(`Make sure ${g} < ${A} <= ${Q}.${I ? " " + I : ""}`), this.name = "RangeError";
}
}
class bs extends SE {
constructor(A) {
super(`Make sure ${A} >= 0.`), this.name = "NonNegativeError";
}
}
class Ws extends SE {
constructor(A, g) {
super(`Make sure ${A} to be number represented in string.${g ? " Got " + g : ""}`), this.name = "InvalidNumberStringError";
}
}
class Zs extends SE {
constructor(A, g, Q) {
super(`${A} requires ${g} to be ${Q}.`), this.name = "DependencyError";
}
}
class go extends Error {
constructor() {
super("WebGPU is not supported in your current environment, but it is necessary to run the WebLLM engine. Please make sure that your browser supports WebGPU and that it is enabled in your browser settings. You can also consult your browser's compatibility chart to see if it supports WebGPU. For more information about WebGPU support in your browser, visit https://webgpureport.org/"), this.name = "WebGPUNotAvailableError";
}
}
class Ki extends Error {
constructor(A) {
super(`Model not loaded before trying to complete ${A}. Please ensure you have called MLCEngine.reload(model) to load the model before initiating APIs, or initialize your engine using CreateMLCEngine() with a valid model configuration.`), this.name = "ModelNotLoadedError";
}
}
class qo extends Error {
constructor(A) {
super(A), this.name = "MessageOrderError";
}
}
class ri extends Error {
constructor() {
super("System prompt should always be the first message in `messages`."), this.name = "SystemMessageOrderError";
}
}
class vs extends Error {
constructor(A) {
super(`${A} should have string content.`), this.name = "ContentTypeError";
}
}
class Vs extends Error {
constructor(A) {
super(`Unsupported role of message: ${A}`), this.name = "UnsupportedRoleError";
}
}
class Xs extends Error {
constructor(A, g, Q) {
super(`The model loaded is not of type ModelType.VLM (vision-language model). Therefore, user message only supports string content, but received: ${Q}
Loaded modelId: ${A}, modelType: ${g}`), this.name = "UserMessageContentErrorForNonVLM";
}
}
class zs extends Error {
constructor(A, g) {
super(`prefillChunkSize needs to be greater than imageEmbedSize because a single image's prefill cannot be chunked. Got prefillChunkSize: ${A}, imageEmbedSize: ${g}`), this.name = "PrefillChunkSizeSmallerThanImageError";
}
}
class _s extends Error {
constructor() {
super("Received image input but model does not have kernel image_embed. Make sure to only pass in image to a vision model."), this.name = "CannotFindImageEmbedError";
}
}
class $s extends Error {
constructor(A) {
super(`Currently do not support field image_url.detail, but received: ${A}`), this.name = "UnsupportedDetailError";
}
}
class Aa extends Error {
constructor(A) {
super(`image_url.url should start with "data:image" for base64, or with "http", but got: ${A}`), this.name = "UnsupportedImageURLError";
}
}
class Ri extends Error {
constructor() {
super("Each message can have at most one text contentPart, but received more than 1."), this.name = "MultipleTextContentError";
}
}
class ga extends Error {
constructor(A, g) {
super(`Internal error: error encountered when parsing outputMessage for function calling. Got outputMessage: ${A}
Got error: ${g}`), this.name = "ToolCallOutputParseError";
}
}
class Qa extends Error {
constructor(A) {
super(`Internal error: expect output of function calling to be an ${A}`), this.name = "ToolCallOutputInvalidTypeError";
}
}
class Ba extends Error {
constructor(A, g) {
super(`Expect generated tool call to have fields ${A.map((Q) => `"\`${Q}\`"`).join(", ")}, but got object: ${JSON.stringify(g)}`), this.name = "JSONFieldError";
}
}
class Ia extends Error {
constructor() {
super("Configuration not initialized. Ensure you have called `reload()` function first."), this.name = "ConfigurationNotInitializedError";
}
}
class Ca extends Error {
constructor(A) {
super(`Missing \`model_lib\` for the model with ID "${A}". Please ensure that \`model_lib\` is provided in \`model_list\` for each model. This URL is essential for downloading the WASM library necessary to run the model.`), this.name = "MissingModelError";
}
}
class yi extends Error {
constructor(A) {
super(`This model requires feature ${A}, which is not yet supported by this browser.`), this.name = "FeatureSupportError";
}
}
class jo extends Error {
constructor(A, g) {
super(`The following fields in ${g} are not yet supported:
` + A.join(", ")), this.name = "UnsupportedFieldsError";
}
}
class Ea extends yi {
constructor() {
super('This model requires WebGPU extension shader-f16, which is not enabled in this browser. You can try to launch Chrome Canary in command line with flag "--enable-dawn-features=allow_unsafe_apis".'), this.name = "ShaderF16SupportError";
}
}
class Da extends Error {
constructor() {
super("The WebGPU device was lost while loading the model. This issue often occurs due to running out of memory (OOM). To resolve this, try reloading with a model that has fewer parameters or uses a smaller context length."), this.name = "DeviceLostError";
}
}
class oa extends Error {
constructor(A) {
super(`Cannot handle tokenizer files ${A}`), this.name = "UnsupportedTokenizerFilesError";
}
}
class wa extends Error {
constructor(A, g) {
super(`Only one of context_window_size and sliding_window_size can be positive. Got: context_window_size: ${A}, sliding_window_size: ${g}
Consider modifying ModelRecord.overrides to set one of them to -1.`), this.name = "WindowSizeConfigurationError";
}
}
class ia extends Error {
constructor() {
super("Need to specify non-negative attention_sink_size if using sliding window. Consider modifying ModelRecord.overrides. Use `attention_sink_size=0` for default sliding window."), this.name = "AttentionSinkSizeError";
}
}
class sa extends Error {
constructor() {
super(`Need to specify either sliding_window_size or max_window_size.
Consider modifying ModelRecord.overrides to set one of them to positive.`), this.name = "WindowSizeSpecificationError";
}
}
class aa extends Error {
constructor(A, g) {
super(`Prompt tokens exceed context window size: number of prompt tokens: ${A}; context window size: ${g}
Consider shortening the prompt, or increase \`context_window_size\`, or using sliding window via \`sliding_window_size\`.`), this.name = "ContextWindowSizeExceededError";
}
}
class Ji extends Error {
constructor() {
super("When streaming, `n` cannot be > 1."), this.name = "StreamingCountError";
}
}
class Li extends Error {
constructor(A) {
super("`seed` should be an integer, but got " + A), this.name = "SeedTypeError";
}
}
class ea extends Error {
constructor() {
super("JSON schema is only supported with `json_object` response format."), this.name = "InvalidResponseFormatError";
}
}
class sw extends Error {
constructor() {
super("When ResponseFormat.type is `grammar`, ResponseFormat.grammar needs to be specified.\nWhen ResponseFormat.grammar is specified, ResponseFormat.type needs to be grammar."), this.name = "InvalidResponseFormatGrammarError";
}
}
class Ga extends Error {
constructor(A) {
super("When using Hermes-2-Pro function calling via ChatCompletionRequest.tools, cannot specify customized response_format. We will set it for you internally. Currently set to: " + JSON.stringify(A)), this.name = "CustomResponseFormatError";
}
}
class ta extends Error {
constructor(A, g) {
super(`${A} is not supported for ChatCompletionRequest.tools. Currently, models that support function calling are: ${g.join(", ")}`), this.name = "UnsupportedModelIdError";
}
}
class Na extends Error {
constructor() {
super("When using Hermes-2-Pro function calling via ChatCompletionRequest.tools, cannot specify customized system prompt."), this.name = "CustomSystemPromptError";
}
}
class ni extends Error {
constructor() {
super("Only specify stream_options when stream=True."), this.name = "InvalidStreamOptionsError";
}
}
class Ya extends Error {
constructor() {
super("Non-chat text completion API expects KVCache to be empty."), this.name = "TextCompletionExpectsKVEmptyError";
}
}
class ka extends Error {
constructor() {
super("Non-chat text completion API expects isTextCompletion is true, and prompt is defined."), this.name = "TextCompletionConversationExpectsPrompt";
}
}
class QE extends Error {
constructor(A) {
super(`Non-chat text completion API cannot call ${A}.`), this.name = "TextCompletionConversationError";
}
}
class ha extends Error {
constructor() {
super("Embedding in base64 format is currently not supported."), this.name = "EmbeddingUnsupportedEncodingFormatError";
}
}
class Ma extends Error {
constructor(A) {
super(`Trying to run embeddings.create() with ${A}, which does not have ModelRecord.model_type === ModelType.embedding in the model record. Either make sure an embedding model is loaded, or specify the model type in ModelRecord.`), this.name = "EmbeddingUnsupportedModelError";
}
}
class ca extends Error {
constructor(A) {
super(`Embedding should not use sliding window. However, sliding_window_size=${A} is specified in the chat config.`), this.name = "EmbeddingSlidingWindowError";
}
}
class Fa extends Error {
constructor(A, g) {
super(`Embedding currently does not support chunking. Make sure contextWindowSize === prefillChunkSize. Got contextWindowSize=${A}, prefillChunkSize=${g} instead.`), this.name = "EmbeddingChunkingUnsupportedError";
}
}
class Ua extends Error {
constructor(A, g) {
super(`The embedding model you are using only supports up to ${A} context size.However, an input in the batch has size ${g}.`), this.name = "EmbeddingExceedContextWindowSizeError";
}
}
class AD extends Error {
constructor() {
super("Embedding input cannot be empty string or empty token array."), this.name = "EmbeddingInputEmptyError";
}
}
class Ka extends Error {
constructor(A, g) {
super(`Expect chatOpts, if specified, to match the size of modelId. However, got ${A} modelId, but ${g} chatOpts.`), this.name = "ReloadArgumentSizeUnmatchedError";
}
}
class ra extends Error {
constructor(A, g) {
super(`Multiple models are loaded in engine. Please specify the model in ${g}.
Currently loaded models are:
${A}`), this.name = "UnclearModelToUseError";
}
}
class Si extends Error {
constructor(A, g, Q) {
super(`Specified model ${g} for ${Q} is not found in loaded models. Please check if the correct model is loaded/specified. Currently loaded models are:
${A}`), this.name = "SpecifiedModelNotFoundError";
}
}
class aw extends Error {
constructor(A, g, Q) {
super(`${Q} expects model to be loaded with ${g}. However, ${A} is not loaded with this pipeline.`), this.name = "IncorrectPipelineLoadedError";
}
}
class Ra extends Error {
constructor(A) {
super(`Need to make models in modelId passed to reload() need to be unique. If you want to, load copies of the same model, consider making copies of the ModelRecord with different model_id. Received modelId: ${A}`), this.name = "ReloadModelIdNotUniqueError";
}
}
var aB;
(function(B) {
B.user = "user", B.assistant = "assistant", B.tool = "tool";
})(aB || (aB = {}));
const ya = "WARN";
var SI;
(function(B) {
B.system = "{system_message}", B.user = "{user_message}", B.assistant = "{assistant_message}", B.tool = "{tool_message}", B.function = "{function_string}", B.hermes_tools = "{hermes_tools}";
})(SI || (SI = {}));
function ew(B) {
function A(g) {
return g != null;
}
if (B.frequency_penalty && (B.frequency_penalty < -2 || B.frequency_penalty > 2))
throw new nI("frequency_penalty", -2, 2);
if (B.presence_penalty && (B.presence_penalty < -2 || B.presence_penalty > 2))
throw new nI("presence_penalty", -2, 2);
if (A(B.repetition_penalty) && B.repetition_penalty <= 0)
throw new wI("repetition_penalty", 0);
if (A(B.max_tokens) && B.max_tokens <= 0)
throw new wI("max_tokens", 0);
if (A(B.top_p) && B.top_p <= 0 || B.top_p > 1)
throw new nI("top_p", 0, 1);
if (A(B.temperature) && B.temperature < 0)
throw new bs("temperature");
if (A(B.frequency_penalty) && !A(B.presence_penalty) && (B.presence_penalty = 0, Og.warn("Only frequency_penalty is set; we default presence_penaty to 0.")), A(B.presence_penalty) && !A(B.frequency_penalty) && (B.frequency_penalty = 0, Og.warn("Only presence_penalty is set; we default frequency_penalty to 0.")), A(B.logit_bias))
for (const g in B.logit_bias) {
const Q = B.logit_bias[g];
if (Q > 100 || Q < -100)
throw new nI("logit_bias", -100, 100, "Got " + Q + " for tokenID " + g);
if (isNaN(parseInt(g)))
throw new Ws("logit_bias's keys", g);
}
if (A(B.top_logprobs)) {
if (!B.logprobs)
throw new Zs("top_logprobs", "logprobs", !0);
if (B.top_logprobs < 0 || B.top_logprobs > 5)
throw new nI("top_logprobs", 0, 5, "Got " + B.top_logprobs);
}
B.logprobs && (A(B.top_logprobs) || (B.top_logprobs = 0));
}
var rQ;
(function(B) {
B[B.LLM = 0] = "LLM", B[B.embedding = 1] = "embedding", B[B.VLM = 2] = "VLM";
})(rQ || (rQ = {}));
const CA = "v0_2_48", EA = "https://raw.githubusercontent.com/mlc-ai/binary-mlc-llm-libs/main/web-llm-models/", Gw = [
"Hermes-2-Pro-Llama-3-8B-q4f16_1-MLC",
"Hermes-2-Pro-Llama-3-8B-q4f32_1-MLC",
"Hermes-2-Pro-Mistral-7B-q4f16_1-MLC",
"Hermes-3-Llama-3.1-8B-q4f32_1-MLC",
"Hermes-3-Llama-3.1-8B-q4f16_1-MLC"
], Ja = {
useIndexedDBCache: !1,
model_list: [
{
model: "https://huggingface.co/mlc-ai/Llama-3.2-1B-Instruct-q4f32_1-MLC",
model_id: "Llama-3.2-1B-Instruct-q4f32_1-MLC",
model_lib: EA + CA + "/Llama-3.2-1B-Instruct-q4f32_1-ctx4k_cs1k-webgpu.wasm",
vram_required_MB: 1128.82,
low_resource_required: !0,
overrides: {
context_window_size: 4096
}
},
{
model: "https://huggingface.co/mlc-ai/Llama-3.2-1B-Instruct-q4f16_1-MLC",
model_id: "Llama-3.2-1B-Instruct-q4f16_1-MLC",
model_lib: EA + CA + "/Llama-3.2-1B-Instruct-q4f16_1-ctx4k_cs1k-webgpu.wasm",
vram_required_MB: 879.04,
low_resource_required: !0,
overrides: {
context_window_size: 4096
}
},
{
model: "https://huggingface.co/mlc-ai/Llama-3.2-1B-Instruct-q0f32-MLC",
model_id: "Llama-3.2-1B-Instruct-q0f32-MLC",
model_lib: EA + CA + "/Llama-3.2-1B-Instruct-q0f32-ctx4k_cs1k-webgpu.wasm",
vram_required_MB: 5106.26,
low_resource_required: !0,
overrides: {
context_window_size: 4096
}
},
{
model: "https://huggingface.co/mlc-ai/Llama-3.2-1B-Instruct-q0f16-MLC",
model_id: "Llama-3.2-1B-Instruct-q0f16-MLC",
model_lib: EA + CA + "/Llama-3.2-1B-Instruct-q0f16-ctx4k_cs1k-webgpu.wasm",
vram_required_MB: 2573.13,
low_resource_required: !0,
overrides: {
context_window_size: 4096
}
},
{
model: "https://huggingface.co/mlc-ai/Llama-3.2-3B-Instruct-q4f32_1-MLC",
model_id: "Llama-3.2-3B-Instruct-q4f32_1-MLC",
model_lib: EA + CA + "/Llama-3.2-3B-Instruct-q4f32_1-ctx4k_cs1k-webgpu.wasm",
vram_required_MB: 2951.51,
low_resource_required: !0,
overrides: {
context_window_size: 4096
}
},
{
model: "https://huggingface.co/mlc-ai/Llama-3.2-3B-Instruct-q4f16_1-MLC",
model_id: "Llama-3.2-3B-Instruct-q4f16_1-MLC",
model_lib: EA + CA + "/Llama-3.2-3B-Instruct-q4f16_1-ctx4k_cs1k-webgpu.wasm",
vram_required_MB: 2263.69,
low_resource_required: !0,
overrides: {
context_window_size: 4096
}
},
// Llama-3.1
{
model: "https://huggingface.co/mlc-ai/Llama-3.1-8B-Instruct-q4f32_1-MLC",
model_id: "Llama-3.1-8B-Instruct-q4f32_1-MLC-1k",
model_lib: EA + CA + "/Llama-3_1-8B-Instruct-q4f32_1-ctx4k_cs1k-webgpu.wasm",
vram_required_MB: 5295.7,
low_resource_required: !0,
overrides: {
context_window_size: 1024
}
},
{
model: "https://huggingface.co/mlc-ai/Llama-3.1-8B-Instruct-q4f16_1-MLC",
model_id: "Llama-3.1-8B-Instruct-q4f16_1-MLC-1k",
model_lib: EA + CA + "/Llama-3_1-8B-Instruct-q4f16_1-ctx4k_cs1k-webgpu.wasm",
vram_required_MB: 4598.34,
low_resource_required: !0,
overrides: {
context_window_size: 1024
}
},
{
model: "https://huggingface.co/mlc-ai/Llama-3.1-8B-Instruct-q4f32_1-MLC",
model_id: "Llama-3.1-8B-Instruct-q4f32_1-MLC",
model_lib: EA + CA + "/Llama-3_1-8B-Instruct-q4f32_1-ctx4k_cs1k-webgpu.wasm",
vram_required_MB: 6101.01,
low_resource_required: !1,
overrides: {
context_window_size: 4096
}
},
{
model: "https://huggingface.co/mlc-ai/Llama-3.1-8B-Instruct-q4f16_1-MLC",
model_id: "Llama-3.1-8B-Instruct-q4f16_1-MLC",
model_lib: EA + CA + "/Llama-3_1-8B-Instruct-q4f16_1-ctx4k_cs1k-webgpu.wasm",
vram_required_MB: 5001,
low_resource_required: !1,
overrides: {
context_window_size: 4096
}
},
// Hermes-2
{
model: "https://huggingface.co/mlc-ai/Hermes-2-Theta-Llama-3-8B-q4f16_1-MLC",
model_id: "Hermes-2-Theta-Llama-3-8B-q4f16_1-MLC",
model_lib: EA + CA + "/Llama-3-8B-Instruct-q4f16_1-ctx4k_cs1k-webgpu.wasm",
vram_required_MB: 4976.13,
low_resource_required: !1,
overrides: {
context_window_size: 4096
}
},
{
model: "https://huggingface.co/mlc-ai/Hermes-2-Theta-Llama-3-8B-q4f32_1-MLC",
model_id: "Hermes-2-Theta-Llama-3-8B-q4f32_1-MLC",
model_lib: EA + CA + "/Llama-3-8B-Instruct-q4f32_1-ctx4k_cs1k-webgpu.wasm",
vram_required_MB: 6051.27,
low_resource_required: !1,
overrides: {
context_window_size: 4096
}
},
{
model: "https://huggingface.co/mlc-ai/Hermes-2-Pro-Llama-3-8B-q4f16_1-MLC",
model_id: "Hermes-2-Pro-Llama-3-8B-q4f16_1-MLC",
model_lib: EA + CA + "/Llama-3-8B-Instruct-q4f16_1-ctx4k_cs1k-webgpu.wasm",
vram_required_MB: 4976.13,
low_resource_required: !1,
overrides: {
context_window_size: 4096
}
},
{
model: "https://huggingface.co/mlc-ai/Hermes-2-Pro-Llama-3-8B-q4f32_1-MLC",
model_id: "Hermes-2-Pro-Llama-3-8B-q4f32_1-MLC",
model_lib: EA + CA + "/Llama-3-8B-Instruct-q4f32_1-ctx4k_cs1k-webgpu.wasm",
vram_required_MB: 6051.27,
low_resource_required: !1,
overrides: {
context_window_size: 4096
}
},
{
model: "https://huggingface.co/mlc-ai/Hermes-3-Llama-3.1-8B-q4f32_1-MLC",
model_id: "Hermes-3-Llama-3.1-8B-q4f32_1-MLC",
model_lib: EA + CA + "/Llama-3_1-8B-Instruct-q4f32_1-ctx4k_cs1k-webgpu.wasm",
vram_required_MB: 5779.27,
low_resource_required: !1,
overrides: {
context_window_size: 4096
}
},
{
model: "https://huggingface.co/mlc-ai/Hermes-3-Llama-3.1-8B-q4f16_1-MLC",
model_id: "Hermes-3-Llama-3.1-8B-q4f16_1-MLC",
model_lib: EA + CA + "/Llama-3_1-8B-Instruct-q4f16_1-ctx4k_cs1k-webgpu.wasm",
vram_required_MB: 4876.13,
low_resource_required: !1,
overrides: {
context_window_size: 4096
}
},
{
model: "https://huggingface.co/mlc-ai/Hermes-2-Pro-Mistral-7B-q4f16_1-MLC",
model_id: "Hermes-2-Pro-Mistral-7B-q4f16_1-MLC",
model_lib: EA + CA + "/Mistral-7B-Instruct-v0.3-q4f16_1-ctx4k_cs1k-webgpu.wasm",
vram_required_MB: 4033.28,
low_resource_required: !1,
required_features: ["shader-f16"],
overrides: {
context_window_size: 4096,
sliding_window_size: -1
}
},
// Phi3.5-mini-instruct
{
model: "https://huggingface.co/mlc-ai/Phi-3.5-mini-instruct-q4f16_1-MLC",
model_id: "Phi-3.5-mini-instruct-q4f16_1-MLC",
model_lib: EA + CA + "/Phi-3.5-mini-instruct-q4f16_1-ctx4k_cs1k-webgpu.wasm",
vram_required_MB: 3672.07,
low_resource_required: !1,
overrides: {
context_window_size: 4096
}
},
{
model: "https://huggingface.co/mlc-ai/Phi-3.5-mini-instruct-q4f32_1-MLC",
model_id: "Phi-3.5-mini-instruct-q4f32_1-MLC",
model_lib: EA + CA + "/Phi-3.5-mini-instruct-q4f32_1-ctx4k_cs1k-webgpu.wasm",
vram_required_MB: 5483.12,
low_resource_required: !1,
overrides: {
context_window_size: 4096
}
},
{
model: "https://huggingface.co/mlc-ai/Phi-3.5-mini-instruct-q4f16_1-MLC",
model_id: "Phi-3.5-mini-instruct-q4f16_1-MLC-1k",
model_lib: EA + CA + "/Phi-3.5-mini-instruct-q4f16_1-ctx4k_cs1k-webgpu.wasm",
vram_required_MB: 2520.07,
low_resource_required: !0,
overrides: {
context_window_size: 1024
}
},
{
model: "https://huggingface.co/mlc-ai/Phi-3.5-mini-instruct-q4f32_1-MLC",
model_id: "Phi-3.5-mini-instruct-q4f32_1-MLC-1k",
model_lib: EA + CA + "/Phi-3.5-mini-instruct-q4f32_1-ctx4k_cs1k-webgpu.wasm",
vram_required_MB: 3179.12,
low_resource_required: !0,
overrides: {
context_window_size: 1024
}
},
// Phi-3.5-vision-instruct
{
model: "https://huggingface.co/mlc-ai/Phi-3.5-vision-instruct-q4f16_1-MLC",
model_id: "Phi-3.5-vision-instruct-q4f16_1-MLC",
model_lib: EA + CA + "/Phi-3.5-vision-instruct-q4f16_1-ctx4k_cs2k-webgpu.wasm",
vram_required_MB: 3952.18,
low_resource_required: !0,
overrides: {
context_window_size: 4096
},
model_type: rQ.VLM
},
{
model: "https://huggingface.co/mlc-ai/Phi-3.5-vision-instruct-q4f32_1-MLC",
model_id: "Phi-3.5-vision-instruct-q4f32_1-MLC",
model_lib: EA + CA + "/Phi-3.5-vision-instruct-q4f32_1-ctx4k_cs2k-webgpu.wasm",
vram_required_MB: 5879.84,
low_resource_required: !0,
overrides: {
context_window_size: 4096
},
model_type: rQ.VLM
},
// Mistral variants
{
model: "https://huggingface.co/mlc-ai/Mistral-7B-Instruct-v0.3-q4f16_1-MLC",
model_id: "Mistral-7B-Instruct-v0.3-q4f16_1-MLC",
model_lib: EA + CA + "/Mistral-7B-Instruct-v0.3-q4f16_1-ctx4k_cs1k-webgpu.wasm",
vram_required_MB: 4573.39,
low_resource_required: !1,
required_features: ["shader-f16"],
overrides: {
context_window_size: 4096,
sliding_window_size: -1
}
},
{
model: "https://huggingface.co/mlc-ai/Mistral-7B-Instruct-v0.3-q4f32_1-MLC",
model_id: "Mistral-7B-Instruct-v0.3-q4f32_1-MLC",
model_lib: EA + CA + "/Mistral-7B-Instruct-v0.3-q4f32_1-ctx4k_cs1k-webgpu.wasm",
vram_required_MB: 5619.27,
low_resource_required: !1,
overrides: {
context_window_size: 4096,
sliding_window_size: -1
}
},
{
model: "https://huggingface.co/mlc-ai/Mistral-7B-Instruct-v0.2-q4f16_1-MLC",
model_id: "Mistral-7B-Instruct-v0.2-q4f16_1-MLC",
model_lib: EA + CA + "/Mistral-7B-Instruct-v0.3-q4f16_1-ctx4k_cs1k-webgpu.wasm",
vram_required_MB: 4573.39,
low_resource_required: !1,
required_features: ["shader-f16"],
overrides: {
context_window_size: 4096,
sliding_window_size: -1
}
},
{
model: "https://huggingface.co/mlc-ai/OpenHermes-2.5-Mistral-7B-q4f16_1-MLC",
model_id: "OpenHermes-2.5-Mistral-7B-q4f16_1-MLC",
model_lib: EA + CA + "/Mistral-7B-Instruct-v0.3-q4f16_1-ctx4k_cs1k-webgpu.wasm",
vram_required_MB: 4573.39,
low_resource_required: !1,
required_features: ["shader-f16"],
overrides: {
context_window_size: 4096,
sliding_window_size: -1
}
},
{
model: "https://huggingface.co/mlc-ai/NeuralHermes-2.5-Mistral-7B-q4f16_1-MLC",
model_id: "NeuralHermes-2.5-Mistral-7B-q4f16_1-MLC",
model_lib: EA + CA + "/Mistral-7B-Instruct-v0.3-q4f16_1-ctx4k_cs1k-webgpu.wasm",
vram_required_MB: 4573.39,
low_resource_required: !1,
required_features: ["shader-f16"],
overrides: {
context_window_size: 4096,
sliding_window_size: -1
}
},
{
model: "https://huggingface.co/mlc-ai/WizardMath-7B-V1.1-q4f16_1-MLC",
model_id: "WizardMath-7B-V1.1-q4f16_1-MLC",
model_lib: EA + CA + "/Mistral-7B-Instruct-v0.3-q4f16_1-ctx4k_cs1k-webgpu.wasm",
vram_required_MB: 4573.39,
low_resource_required: !1,
required_features: ["shader-f16"],
overrides: {
context_window_size: 4096,
sliding_window_size: -1
}
},
// SmolLM2
{
model: "https://huggingface.co/mlc-ai/SmolLM2-1.7B-Instruct-q4f16_1-MLC",
model_id: "SmolLM2-1.7B-Instruct-q4f16_1-MLC",
model_lib: EA + CA + "/SmolLM2-1.7B-Instruct-q4f16_1-ctx4k_cs1k-webgpu.wasm",
vram_required_MB: 1774.19,
low_resource_required: !0,
required_features: ["shader-f16"],
overrides: {
context_window_size: 4096
}
},
{
model: "https://huggingface.co/mlc-ai/SmolLM2-1.7B-Instruct-q4f32_1-MLC",
model_id: "SmolLM2-1.7B-Instruct-q4f32_1-MLC",
model_lib: EA + CA + "/SmolLM2-1.7B-Instruct-q4f32_1-ctx4k_cs1k-webgpu.wasm",
vram_required_MB: 2692.38,
low_resource_required: !0,
overrides: {
context_window_size: 4096
}
},
{
model: "https://huggingface.co/mlc-ai/SmolLM2-360M-Instruct-q0f16-MLC",
model_id: "SmolLM2-360M-Instruct-q0f16-MLC",
model_lib: EA + CA + "/SmolLM2-360M-Instruct-q0f16-ctx4k_cs1k-webgpu.wasm",
vram_required_MB: 871.99,
low_resource_required: !0,
required_features: ["shader-f16"],
overrides: {
context_window_size: 4096
}
},
{
model: "https://huggingface.co/mlc-ai/SmolLM2-360M-Instruct-q0f32-MLC",
model_id: "SmolLM2-360M-Instruct-q0f32-MLC",
model_lib: EA + CA + "/SmolLM2-360M-Instruct-q0f32-ctx4k_cs1k-webgpu.wasm",
vram_required_MB: 1743.99,
low_resource_required: !0,
overrides: {
context_window_size: 4096
}
},
{
model: "https://huggingface.co/mlc-ai/SmolLM2-360M-Instruct-q4f16_1-MLC",
model_id: "SmolLM2-360M-Instruct-q4f16_1-MLC",
model_lib: EA + CA + "/SmolLM2-360M-Instruct-q4f16_1-ctx4k_cs1k-webgpu.wasm",
vram_required_MB: 376.06,
low_resource_required: !0,
required_features: ["shader-f16"],
overrides: {
context_window_size: 4096
}
},
{
model: "https://huggingface.co/mlc-ai/SmolLM2-360M-Instruct-q4f32_1-MLC",
model_id: "SmolLM2-360M-Instruct-q4f32_1-MLC",
model_lib: EA + CA + "/SmolLM2-360M-Instruct-q4f32_1-ctx4k_cs1k-webgpu.wasm",
vram_required_MB: 579.61,
low_resource_required: !0,
overrides: {
context_window_size: 4096
}
},
{
model: "https://huggingface.co/mlc-ai/SmolLM2-135M-Instruct-q0f16-MLC",
model_id: "SmolLM2-135M-Instruct-q0f16-MLC",
model_lib: EA + CA + "/SmolLM2-135M-Instruct-q0f16-ctx4k_cs1k-webgpu.wasm",
vram_required_MB: 359.69,
low_resource_required: !0,
required_features: ["shader-f16"],
overrides: {
context_window_size: 4096
}
},
{
model: "https://huggingface.co/mlc-ai/SmolLM2-135M-Instruct-q0f32-MLC",
model_id: "SmolLM2-135M-Instruct-q0f32-MLC",
model_lib: EA + CA + "/SmolLM2-135M-Instruct-q0f32-ctx4k_cs1k-webgpu.wasm",
vram_required_MB: 719.38,
low_resource_required: !0,
overrides: {
context_window_size: 4096
}
},
// Gemma2
{
model: "https://huggingface.co/mlc-ai/gemma-2-2b-it-q4f16_1-MLC",
model_id: "gemma-2-2b-it-q4f16_1-MLC",
model_lib: EA + CA + "/gemma-2-2b-it-q4f16_1-ctx4k_cs1k-webgpu.wasm",
vram_required_MB: 1895.3,
low_resource_required: !1,
required_features: ["shader-f16"],
overrides: {
context_window_size: 4096
}
},
{
model: "https://huggingface.co/mlc-ai/gemma-2-2b-it-q4f32_1-MLC",
model_id: "gemma-2-2b-it-q4f32_1-MLC",
model_lib: EA + CA + "/gemma-2-2b-it-q4f32_1-ctx4k_cs1k-webgpu.wasm",
vram_required_MB: 2508.75,
low_resource_required: !1,
overrides: {
context_window_size: 4096
}
},
{
model: "https://huggingface.co/mlc-ai/gemma-2-2b-it-q4f16_1-MLC",
model_id: "gemma-2-2b-it-q4f16_1-MLC-1k",
model_lib: EA + CA + "/gemma-2-2b-it-q4f16_1-ctx4k_cs1k-webgpu.wasm",
vram_required_MB: 1583.3,
low_resource_required: !0,
required_features: ["shader-f16"],
overrides: {
context_window_size: 1024
}
},
{
model: "https://huggingface.co/mlc-ai/gemma-2-2b-it-q4f32_1-MLC",
model_id: "gemma-2-2b-it-q4f32_1-MLC-1k",
model_lib: EA + CA + "/gemma-2-2b-it-q4f32_1-ctx4k_cs1k-webgpu.wasm",
vram_required_MB: 1884.75,
low_resource_required: !0,
overrides: {
context_window_size: 1024
}
},
{
model: "https://huggingface.co/mlc-ai/gemma-2-9b-it-q4f16_1-MLC",
model_id: "gemma-2-9b-it-q4f16_1-MLC",
model_lib: EA + CA + "/gemma-2-9b-it-q4f16_1-ctx4k_cs1k-webgpu.wasm",
vram_required_MB: 6422.01,
low_resource_required: !1,
required_features: ["shader-f16"],
overrides: {
context_window_size: 4096
}
},
{
model: "https://huggingface.co/mlc-ai/gemma-2-9b-it-q4f32_1-MLC",
model_id: "gemma-2-9b-it-q4f32_1-MLC",
model_lib: EA + CA + "/gemma-2-9b-it-q4f32_1-ctx4k_cs1k-webgpu.wasm",
vram_required_MB: 8383.33,
low_resource_required: !1,
overrides: {
context_window_size: 4096
}
},
// Gemma2-2b-jpn
{
model: "https://huggingface.co/mlc-ai/gemma-2-2b-jpn-it-q4f16_1-MLC",
model_id: "gemma-2-2b-jpn-it-q4f16_1-MLC",
model_lib: EA + CA + "/gemma-2-2b-jpn-it-q4f16_1-ctx4k_cs1k-webgpu.wasm",
vram_required_MB: 1895.3,
low_resource_required: !0,
required_features: ["shader-f16"],
overrides: {
context_window_size: 4096
}
},
{
model: "https://huggingface.co/mlc-ai/gemma-2-2b-jpn-it-q4f32_1-MLC",
model_id: "gemma-2-2b-jpn-it-q4f32_1-MLC",
model_lib: EA + CA + "/gemma-2-2b-jpn-it-q4f32_1-ctx4k_cs1k-webgpu.wasm",
vram_required_MB: 2508.75,
low_resource_required: !0,
overrides: {
context_window_size: 4096
}
},
// Qwen-2
{
model: "https://huggingface.co/mlc-ai/Qwen2.5-0.5B-Instruct-q4f16_1-MLC",
model_id: "Qwen2.5-0.5B-Instruct-q4f16_1-MLC",
model_lib: EA + CA + "/Qwen2-0.5B-Instruct-q4f16_1-ctx4k_cs1k-webgpu.wasm",
low_resource_required: !0,
vram_required_MB: 944.62,
overrides: {
context_window_size: 4096
}
},
{
model: "https://huggingface.co/mlc-ai/Qwen2.5-0.5B-Instruct-q4f32_1-MLC",
model_id: "Qwen2.5-0.5B-Instruct-q4f32_1-MLC",
model_lib: EA + CA + "/Qwen2-0.5B-Instruct-q4f32_1-ctx4k_cs1k-webgpu.wasm",
low_resource_required: !0,
vram_required_MB: 1060.2,
overrides: {
context_window_size: 4096
}
},
{
model: "https://huggingface.co/mlc-ai/Qwen2.5-0.5B-Instruct-q0f16-MLC",
model_id: "Qwen2.5-0.5B-Instruct-q0f16-MLC",
model_lib: EA + CA + "/Qwen2-0.5B-Instruct-q0f16-ctx4k_cs1k-webgpu.wasm",
low_resource_required: !0,
vram_required_MB: 1624.12,
overrides: {
context_window_size: 4096
}
},
{
model: "https://huggingface.co/mlc-ai/Qwen2.5-0.5B-Instruct-q0f32-MLC",
model_id: "Qwen2.5-0.5B-Instruct-q0f32-MLC",
model_lib: EA + CA + "/Qwen2-0.5B-Instruct-q0f32-ctx4k_cs1k-webgpu.wasm",
low_resource_required: !0,
vram_required_MB: 2654.75,
overrides: {
context_window_size: 4096
}
},
{
model: "https://huggingface.co/mlc-ai/Qwen2.5-1.5B-Instruct-q4f16_1-MLC",
model_id: "Qwen2.5-1.5B-Instruct-q4f16_1-MLC",
model_lib: EA + CA + "/Qwen2-1.5B-Instruct-q4f16_1-ctx4k_cs1k-webgpu.wasm",
low_resource_required: !0,
vram_required_MB: 1629.75,
overrides: {
context_window_size: 4096
}
},
{
model: "https://huggingface.co/mlc-ai/Qwen2.5-1.5B-Instruct-q4f32_1-MLC",
model_id: "Qwen2.5-1.5B-Instruct-q4f32_1-MLC",
model_lib: EA + CA + "/Qwen2-1.5B-Instruct-q4f32_1-ctx4k_cs1k-webgpu.wasm",
low_resource_required: !0,
vram_required_MB: 1888.97,
overrides: {
context_window_size: 4096
}
},
{
model: "https://huggingface.co/mlc-ai/Qwen2.5-3B-Instruct-q4f16_1-MLC",
model_id: "Qwen2.5-3B-Instruct-q4f16_1-MLC",
model_lib: EA + CA + "/Qwen2.5-3B-Instruct-q4f16_1-ctx4k_cs1k-webgpu.wasm",
low_resource_required: !0,
vram_required_MB: 2504.76,
overrides: {
context_window_size: 4096
}
},
{
model: "https://huggingface.co/mlc-ai/Qwen2.5-3B-Instruct-q4f32_1-MLC",
model_id: "Qwen2.5-3B-Instruct-q4f32_1-MLC",
model_lib: EA + CA + "/Qwen2.5-3B-Instruct-q4f32_1-ctx4k_cs1k-webgpu.wasm",
low_resource_required: !0,
vram_required_MB: 2893.64,
overrides: {
context_window_size: 4096
}
},
{
model: "https://huggingface.co/mlc-ai/Qwen2.5-7B-Instruct-q4f16_1-MLC",
model_id: "Qwen2.5-7B-Instruct-q4f16_1-MLC",
model_lib: EA + CA + "/Qwen2-7B-Instruct-q4f16_1-ctx4k_cs1k-webgpu.wasm",
low_resource_required: !1,
vram_required_MB: 5106.67,
overrides: {
context_window_size: 4096
}
},
{
model: "https://huggingface.co/mlc-ai/Qwen2.5-7B-Instruct-q4f32_1-MLC",
model_id: "Qwen2.5-7B-Instruct-q4f32_1-MLC",
model_lib: EA + CA + "/Qwen2-7B-Instruct-q4f32_1-ctx4k_cs1k-webgpu.wasm",
low_resource_required: !1,
vram_required_MB: 5900.09,
overrides: {
context_window_size: 4096
}
},
// Qwen2.5-Coder
{
model: "https://huggingface.co/mlc-ai/Qwen2.5-Coder-0.5B-Instruct-q4f16_1-MLC",
model_id: "Qwen2.5-Coder-0.5B-Instruct-q4f16_1-MLC",
model_lib: EA + CA + "/Qwen2-0.5B-Instruct-q4f16_1-ctx4k_cs1k-webgpu.wasm",
low_resource_required: !0,
vram_required_MB: 944.62,
overrides: {
context_window_size: 4096
}
},
{
model: "https://huggingface.co/mlc-ai/Qwen2.5-Coder-0.5B-Instruct-q4f32_1-MLC",
model_id: "Qwen2.5-Coder-0.5B-Instruct-q4f32_1-MLC",
model_lib: EA + CA + "/Qwen2-0.5B-Instruct-q4f32_1-ctx4k_cs1k-webgpu.wasm",
low_resource_required: !0,
vram_required_MB: 1060.2,
overrides: {
context_window_size: 4096
}
},
{
model: "https://huggingface.co/mlc-ai/Qwen2.5-Coder-0.5B-Instruct-q0f16-MLC",
model_id: "Qwen2.5-Coder-0.5B-Instruct-q0f16-MLC",
model_lib: EA + CA + "/Qwen2-0.5B-Instruct-q0f16-ctx4k_cs1k-webgpu.wasm",
low_resource_required: !0,
vram_required_MB: 1624.12,
overrides: {
context_window_size: 4096
}
},
{
model: "https://huggingface.co/mlc-ai/Qwen2.5-Coder-0.5B-Instruct-q0f32-MLC",
model_id: "Qwen2.5-Coder-0.5B-Instruct-q0f32-MLC",
model_lib: EA + CA + "/Qwen2-0.5B-Instruct-q0f32-ctx4k_cs1k-webgpu.wasm",
low_resource_required: !0,
vram_required_MB: 2654.75,
overrides: {
context_window_size: 4096
}
},
{
model: "https://huggingface.co/mlc-ai/Qwen2.5-Coder-1.5B-Instruct-q4f16_1-MLC",
model_id: "Qwen2.5-Coder-1.5B-Instruct-q4f16_1-MLC",
model_lib: EA + CA + "/Qwen2-1.5B-Instruct-q4f16_1-ctx4k_cs1k-webgpu.wasm",
low_resource_required: !1,
vram_required_MB: 1629.75,
overrides: {
context_window_size: 4096
}
},
{
model: "https://huggingface.co/mlc-ai/Qwen2.5-Coder-1.5B-Instruct-q4f32_1-MLC",
model_id: "Qwen2.5-Coder-1.5B-Instruct-q4f32_1-MLC",
model_lib: EA + CA + "/Qwen2-1.5B-Instruct-q4f32_1-ctx4k_cs1k-webgpu.wasm",
low_resource_required: !1,
vram_required_MB: 1888.97,
overrides: {
context_window_size: 4096
}
},
{
model: "https://huggingface.co/mlc-ai/Qwen2.5-Coder-3B-Instruct-q4f16_1-MLC",
model_id: "Qwen2.5-Coder-3B-Instruct-q4f16_1-MLC",
model_lib: EA + CA + "/Qwen2.5-3B-Instruct-q4f16_1-ctx4k_cs1k-webgpu.wasm",
low_resource_required: !0,
vram_required_MB: 2504.76,
overrides: {
context_window_size: 4096
}
},
{
model: "https://huggingface.co/mlc-ai/Qwen2.5-Coder-3B-Instruct-q4f32_1-MLC",
model_id: "Qwen2.5-Coder-3B-Instruct-q4f32_1-MLC",
model_lib: EA + CA + "/Qwen2.5-3B-Instruct-q4f32_1-ctx4k_cs1k-webgpu.wasm",
low_resource_required: !0,
vram_required_MB: 2893.64,
overrides: {
context_window_size: 4096
}
},
{
model: "https://huggingface.co/mlc-ai/Qwen2.5-Coder-7B-Instruct-q4f16_1-MLC",
model_id: "Qwen2.5-Coder-7B-Instruct-q4f16_1-MLC",
model_lib: EA + CA + "/Qwen2-7B-Instruct-q4f16_1-ctx4k_cs1k-webgpu.wasm",
low_resource_required: !1,
vram_required_MB: 5106.67,
overrides: {
context_window_size: 4096
}
},
{
model: "https://huggingface.co/mlc-ai/Qwen2.5-Coder-7B-Instruct-q4f32_1-MLC",
model_id: "Qwen2.5-Coder-7B-Instruct-q4f32_1-MLC",
model_lib: EA + CA + "/Qwen2-7B-Instruct-q4f32_1-ctx4k_cs1k-webgpu.wasm",
low_resource_required: !1,
vram_required_MB: 5900.09,
overrides: {
context_window_size: 4096
}
},
// Qwen2.5-Math
{
model: "https://huggingface.co/mlc-ai/Qwen2.5-Math-1.5B-Instruct-q4f16_1-MLC",
model_id: "Qwen2.5-Math-1.5B-Instruct-q4f16_1-MLC",
model_lib: EA + CA + "/Qwen2-1.5B-Instruct-q4f16_1-ctx4k_cs1k-webgpu.wasm",
low_resource_required: !0,
vram_required_MB: 1629.75,
overrides: {
context_window_size: 4096
}
},
{
model: "https://huggingface.co/mlc-ai/Qwen2.5-Math-1.5B-Instruct-q4f32_1-MLC",
model_id: "Qwen2.5-Math-1.5B-Instruct-q4f32_1-MLC",
model_lib: EA + CA + "/Qwen2-1.5B-Instruct-q4f32_1-ctx4k_cs1k-webgpu.wasm",
low_resource_required: !0,
vram_required_MB: 1888.97,
overrides: {
context_window_size: 4096
}
},
// StableLM-zephyr-1.6B
{
model: "https://huggingface.co/mlc-ai/stablelm-2-zephyr-1_6b-q4f16_1-MLC",
model_id: "stablelm-2-zephyr-1_6b-q4f16_1-MLC",
model_lib: EA + CA + "/stablelm-2-zephyr-1_6b-q4f16_1-ctx4k_cs1k-webgpu.wasm",
vram_required_MB: 2087.66,
low_resource_required: !1,
overrides: {
context_window_size: 4096
}
},
{
model: "https://huggingface.co/mlc-ai/stablelm-2-zephyr-1_6b-q4f32_1-MLC",
model_id: "stablelm-2-zephyr-1_6b-q4f32_1-MLC",
model_lib: EA + CA + "/stablelm-2-zephyr-1_6b-q4f32_1-ctx4k_cs1k-webgpu.wasm",
vram_required_MB: 2999.33,
low_resource_required: !1,
overrides: {
context_window_size: 4096
}
},
{
model: "https://huggingface.co/mlc-ai/stablelm-2-zephyr-1_6b-q4f16_1-MLC",
model_id: "stablelm-2-zephyr-1_6b-q4f16_1-MLC-1k",
model_lib: EA + CA + "/stablelm-2-zephyr-1_6b-q4f16_1-ctx4k_cs1k-webgpu.wasm",
vram_required_MB: 1511.66,
low_resource_required: !0,
overrides: {
context_window_size: 1024
}
},
{
model: "https://huggingface.co/mlc-ai/stablelm-2-zephyr-1_6b-q4f32_1-MLC",
model_id: "stablelm-2-zephyr-1_6b-q4f32_1-MLC-1k",
model_lib: EA + CA + "/stablelm-2-zephyr-1_6b-q4f32_1-ctx4k_cs1k-webgpu.wasm",
vram_required_MB: 1847.33,
low_resource_required: !0,
overrides: {
context_window_size: 1024
}
},
// RedPajama
{
model: "https://huggingface.co/mlc-ai/RedPajama-INCITE-Chat-3B-v1-q4f16_1-MLC",
model_id: "RedPajama-INCITE-Chat-3B-v1-q4f16_1-MLC",
model_lib: EA + CA + "/RedPajama-INCITE-Chat-3B-v1-q4f16_1-ctx2k_cs1k-webgpu.wasm",
vram_required_MB: 2972.09,
low_resource_required: !1,
required_features: ["shader-f16"],
overrides: {
context_window_size: 2048
}
},
{
model: "https://huggingface.co/mlc-ai/RedPajama-INCITE-Chat-3B-v1-q4f32_1-MLC",
model_id: "RedPajama-INCITE-Chat-3B-v1-q4f32_1-MLC",
model_lib: EA + CA + "/RedPajama-INCITE-Chat-3B-v1-q4f32_1-ctx2k_cs1k-webgpu.wasm",
vram_required_MB: 3928.09,
low_resource_required: !1,
overrides: {
context_window_size: 2048
}
},
{
model: "https://huggingface.co/mlc-ai/RedPajama-INCITE-Chat-3B-v1-q4f16_1-MLC",
model_id: "RedPajama-INCITE-Chat-3B-v1-q4f16_1-MLC-1k",
model_lib: EA + CA + "/RedPajama-INCITE-Chat-3B-v1-q4f16_1-ctx2k_cs1k-webgpu.wasm",
vram_required_MB: 2041.09,
low_resource_required: !0,
required_features: ["shader-f16"],
overrides: {
context_window_size: 1024
}
},
{
model: "https://huggingface.co/mlc-ai/RedPajama-INCITE-Chat-3B-v1-q4f32_1-MLC",
model_id: "RedPajama-INCITE-Chat-3B-v1-q4f32_1-MLC-1k",
model_lib: EA + CA + "/RedPajama-INCITE-Chat-3B-v1-q4f32_1-ctx2k_cs1k-webgpu.wasm",
vram_required_MB: 2558.09,
low_resource_required: !0,
overrides: {
context_window_size: 1024
}
},
// TinyLlama v1.0
{
model: "https://huggingface.co/mlc-ai/TinyLlama-1.1B-Chat-v1.0-q4f16_1-MLC",
model_id: "TinyLlama-1.1B-Chat-v1.0-q4f16_1-MLC",
model_lib: EA + CA + "/TinyLlama-1.1B-Chat-v1.0-q4f16_1-ctx2k_cs1k-webgpu.wasm",
vram_required_MB: 697.24,
low_resource_required: !0,
required_features: ["shader-f16"],
overrides: {
context_window_size: 2048
}
},
{
model: "https://huggingface.co/mlc-ai/TinyLlama-1.1B-Chat-v1.0-q4f32_1-MLC",
model_id: "TinyLlama-1.1B-Chat-v1.0-q4f32_1-MLC",
model_lib: EA + CA + "/TinyLlama-1.1B-Chat-v1.0-q4f32_1-ctx2k_cs1k-webgpu.wasm",
vram_required_MB: 839.98,
low_resource_required: !0,
overrides: {
context_window_size: 2048
}
},
{
model: "https://huggingface.co/mlc-ai/TinyLlama-1.1B-Chat-v1.0-q4f16_1-MLC",
model_id: "TinyLlama-1.1B-Chat-v1.0-q4f16_1-MLC-1k",
model_lib: EA + CA + "/TinyLlama-1.1B-Chat-v1.0-q4f16_1-ctx2k_cs1k-webgpu.wasm",
vram_required_MB: 675.24,
low_resource_required: !0,
required_features: ["shader-f16"],
overrides: {
context_window_size: 1024
}
},
{
model: "https://huggingface.co/mlc-ai/TinyLlama-1.1B-Chat-v1.0-q4f32_1-MLC",
model_id: "TinyLlama-1.1B-Chat-v1.0-q4f32_1-MLC-1k",
model_lib: EA + CA + "/TinyLlama-1.1B-Chat-v1.0-q4f32_1-ctx2k_cs1k-webgpu.wasm",
vram_required_MB: 795.98,
low_resource_required: !0,
overrides: {
context_window_size: 1024
}
},
// BELOW ARE MODELS OF OLDER VERSIONS OR NOT AS PRACTICAL
// Llama-3.1 70B
{
model: "https://huggingface.co/mlc-ai/Llama-3.1-70B-Instruct-q3f16_1-MLC",
model_id: "Llama-3.1-70B-Instruct-q3f16_1-MLC",
model_lib: EA + CA + "/Llama-3_1-70B-Instruct-q3f16_1-ctx4k_cs1k-webgpu.wasm",
vram_required_MB: 31153.13,
low_resource_required: !1,
overrides: {
context_window_size: 4096
}
},
// Qwen-2
{
model: "https://huggingface.co/mlc-ai/Qwen2-0.5B-Instruct-q4f16_1-MLC",
model_id: "Qwen2-0.5B-Instruct-q4f16_1-MLC",
model_lib: EA + CA + "/Qwen2-0.5B-Instruct-q4f16_1-ctx4k_cs1k-webgpu.wasm",
low_resource_required: !0,
vram_required_MB: 944.62