feat: Add FP8 dtype support (E4M3FN and E5M2)

[nyo16]

Summary

Add FP8 (8-bit floating point) dtype support for E4M3FN and E5M2 formats. This enables reading and writing FP8 quantized model
weights from HuggingFace models like Qwen3-FP8.

Changes

• Add FP8 dtype constants (F8_E4M3FN, F8_E5M2) and type mapping
• Handle 3-tuple FP8 types {:f, 8, :e4m3fn} and {:f, 8, :e5m2} in:
  • dtype_from_string/1 - Parse "F8_E4M3" from safetensor headers
  • tensor_byte_size/1 - Calculate byte size for FP8 tensors
  • tensor_to_iodata/1 - Serialize FP8 tensors
  • build_tensor/2 - Deserialize FP8 tensors
• Support reading FP8 model files (e.g., Qwen/Qwen3-0.6B-FP8)

Test plan

• Unit tests for FP8 type encoding/decoding
• Integration test reading real FP8 model files
• Verified with Qwen3-0.6B-FP8 model inference

Notes

This is the first PR in a series to enable native FP8 model inference:

1. safetensors (this PR) - FP8 file I/O
2. nx/exla - FP8 type system support
3. bumblebee - FP8 model loading and inference

[josevalim]

Please remove the convo.txt :)

My suggestion is to break this in two. The first one is to add FP8 support, which means E5M2. No need for additional tuples and steps.

then a separate PR adds handling for unknown types. For now, the user should pass a separate function that receives the type and the value and builds the tensors

which types QWEN uses?

[nyo16]

Qwen3 is using: F8_E4M3

[nyo16]

ok I will work to break this down, for bumblebee and Nx i will open the PRs as draft for open more the discussion.

[josevalim]

@nyo16 to make everyone on the same page: elixir-nx/nx#1657 (comment)

I think this PR will be straight-forward once we add e4m3fn to Nx, no need for custom functions :)