tiny-d1-Omni-2B-for-testing

This is a modified, damaged copy of LiquidAI/d1-Omni-2B, made only to test software. Its output is gibberish. Do not use it for anything else.

It keeps the architecture, tokenizer, chat template and calibration of d1-Omni-2B, so it runs the same code paths as the real model in llama.cpp (lfm3-d1-omni decision type: non-causal attention, packed questions, image and audio input), but it is small enough for CI.

GGUF version: ggml-org-dev/tiny-d1-Omni-2B-for-testing-GGUF

Changes from the original model

Sliced from the safetensors with slice_d1omni.py, which follows the slice-model skill of llama.cpp:

  • Text model: 4 of 24 layers (0, 2, 3, 4), hidden size 2048 -> 64, heads 16 -> 2, KV heads 4 -> 1, FFN 7680 -> 128
  • Audio encoder: 3 of 17 layers (0, 2, 3), hidden size 768 -> 192, heads 8 -> 2, KV heads 4 -> 1, FFN 1792 -> 256, downsampler channels 256 -> 32
  • Depthformer: 1 of 3 layers, hidden size 1024 -> 128, heads 32 -> 4, KV heads 8 -> 1, FFN 2816 -> 256
  • Vision encoder: 2 of 14 layers (0, 2), hidden size 1024 -> 128, heads 16 -> 2, KV heads 8 -> 1, FFN 2560 -> 128, projector 4096 -> 128
  • The Python code of the original repo is not included

License

Derived from d1-Omni-2B by Liquid AI, under the LFM Open License v1.0.

Downloads last month
-
Safetensors
Model size
14.8M params
Tensor type
BF16
ยท
Inference Providers NEW
This model isn't deployed by any Inference Provider. ๐Ÿ™‹ Ask for provider support

Model tree for ggml-org-dev/tiny-d1-Omni-2B-for-testing

Finetuned
(3)
this model
Quantizations
1 model