# `EMLXAxon.Qwen3.Loader`
[🔗](https://github.com/elixir-nx/emlx/blob/v0.3.0/emlx_axon/lib/emlx_axon/qwen3/loader.ex#L1)

Loads a `lmstudio-community/Qwen3-*-MLX-4bit` checkpoint from disk into
an `%EMLXAxon.Qwen3.Model.State{}` struct.

Each linear weight is stored as a triplet in the safetensors file:
  - `<name>.weight`  — `:u32`, packed int4 data
  - `<name>.scales`  — `:bf16` or `:f16`
  - `<name>.biases`  — same dtype as scales

The loader constructs an annotated `Nx.Tensor` via
`EMLX.Quantization.quantized_tensor/5` so plain `Nx.dot(act, weight, …)`
dispatches transparently to `EMLX.quantized_matmul` via the backend.

## Usage

    {:ok, state} = Loader.load("~/models/Qwen3-0.6B-MLX-4bit")

# `config`

```elixir
@type config() :: %{
  hidden_size: pos_integer(),
  intermediate_size: pos_integer(),
  num_attention_heads: pos_integer(),
  num_key_value_heads: pos_integer(),
  head_dim: pos_integer(),
  num_hidden_layers: pos_integer(),
  vocab_size: pos_integer(),
  rms_norm_eps: float(),
  rope_theta: float(),
  tie_word_embeddings: boolean()
}
```

# `load`

```elixir
@spec load(Path.t()) :: {:ok, EMLXAxon.Qwen3.Model.State.t()} | {:error, term()}
```

Loads a Qwen3 MLX-4bit checkpoint directory.

Reads `config.json` to extract model dimensions, then loads all
`.safetensors` shards found in the directory.

Returns `{:ok, %State{}}` or `{:error, reason}`.

---

*Consult [api-reference.md](api-reference.md) for complete listing*