Entrit commited on
Commit
daebe74
·
verified ·
1 Parent(s): 107da27

Reproducibility: paper checkpoints no longer hosted

Browse files
Files changed (1) hide show
  1. README.md +9 -7
README.md CHANGED
@@ -27,18 +27,20 @@ We evaluate the codec across three model families (Qwen2.5 0.5B–72B, Llama-3.1
27
 
28
  The balanced ternary representation replaces multiplication with conditional addition, enabling multiply-free inference kernels. Per-layer GEMV benchmarks projected to full-model token generation show a **7.8× speedup** over FP16 cuBLAS on an RTX 4090; these are kernel-only projections, not end-to-end serving throughput, and exclude attention, sampling, and tokenizer overhead. This maps to processing-in-memory architectures where eliminating multiplier circuits is the primary design constraint.
29
 
30
- ## Reproducibility — all artifacts public on HuggingFace
31
 
32
  | What | Where | License |
33
  |---|---|---|
34
  | Codec source | [`Entrit/tritllm-codec`](https://huggingface.co/Entrit/tritllm-codec) | Apache 2.0 |
35
  | Multiply-free CUDA kernel | [`Entrit/tritllm-kernel`](https://huggingface.co/Entrit/tritllm-kernel) | Apache 2.0 |
36
- | Pre-quantized Qwen2.5 (0.5B–72B) at *d* = 1, 2, 3, 4 | [Entrit org](https://huggingface.co/Entrit) | Apache 2.0 |
37
- | Pre-quantized Llama-3.1-8B at *d* = 1, 2, 3, 4 | [Entrit org](https://huggingface.co/Entrit) | Llama 3.1 Community |
38
- | Pre-quantized Mistral-7B-v0.3 at *d* = 1, 2, 3, 4 | [Entrit org](https://huggingface.co/Entrit) | Apache 2.0 |
39
- | Block-wise QAT checkpoint (Qwen2.5-7B *d* = 2) | [`Entrit/Qwen2.5-7B-qat-d2`](https://huggingface.co/Entrit/Qwen2.5-7B-qat-d2) | Apache 2.0 |
40
 
41
- Each checkpoint records a codec fingerprint (model id + revision + codec version + depth-power mapping) so reruns are bit-stable.
 
 
 
 
 
 
42
 
43
  ## Citation
44
 
@@ -54,7 +56,7 @@ Each checkpoint records a codec fingerprint (model id + revision + codec version
54
 
55
  ## License and patent notice
56
 
57
- This paper is distributed under the **Creative Commons Attribution 4.0 International License (CC BY 4.0)**. The codec and kernel source are released under **Apache License 2.0**. Quantized model checkpoints inherit their source-model licenses (Apache 2.0 for Qwen2.5 and Mistral derivatives; Llama 3.1 Community License for Llama-3.1-8B derivatives, distributed with required "Built with Llama" attribution).
58
 
59
  The methods described in this paper are the subject of pending patent applications by Entrit Systems. The released software and weights may be used under their respective licenses for research, evaluation, and academic purposes. Commercial deployment of the underlying methods may require a separate license; contact `eric@entrit.io`.
60
 
 
27
 
28
  The balanced ternary representation replaces multiplication with conditional addition, enabling multiply-free inference kernels. Per-layer GEMV benchmarks projected to full-model token generation show a **7.8× speedup** over FP16 cuBLAS on an RTX 4090; these are kernel-only projections, not end-to-end serving throughput, and exclude attention, sampling, and tokenizer overhead. This maps to processing-in-memory architectures where eliminating multiplier circuits is the primary design constraint.
29
 
30
+ ## Reproducibility
31
 
32
  | What | Where | License |
33
  |---|---|---|
34
  | Codec source | [`Entrit/tritllm-codec`](https://huggingface.co/Entrit/tritllm-codec) | Apache 2.0 |
35
  | Multiply-free CUDA kernel | [`Entrit/tritllm-kernel`](https://huggingface.co/Entrit/tritllm-kernel) | Apache 2.0 |
 
 
 
 
36
 
37
+ The pre-quantized checkpoints used in the paper (Qwen2.5 0.5B–72B, Llama-3.1-8B and Mistral-7B-v0.3 at
38
+ *d* = 1–4, and the block-wise QAT Qwen2.5-7B *d* = 2) are no longer hosted. The codec needs no
39
+ calibration data and records a fingerprint (model id + revision + codec version + depth-power mapping)
40
+ in every checkpoint it writes, so rebuilt checkpoints match the paper's.
41
+
42
+ This paper is Entrit's first-generation codec. The current models use the newer d2 codec; see the
43
+ [Entrit page](https://huggingface.co/Entrit).
44
 
45
  ## Citation
46
 
 
56
 
57
  ## License and patent notice
58
 
59
+ This paper is distributed under the **Creative Commons Attribution 4.0 International License (CC BY 4.0)**. The codec and kernel source are released under **Apache License 2.0**. Checkpoints you build with the codec inherit their source-model licenses (Apache 2.0 for Qwen2.5 and Mistral; the Llama 3.1 Community License, with its "Built with Llama" attribution, for Llama-3.1-8B).
60
 
61
  The methods described in this paper are the subject of pending patent applications by Entrit Systems. The released software and weights may be used under their respective licenses for research, evaluation, and academic purposes. Commercial deployment of the underlying methods may require a separate license; contact `eric@entrit.io`.
62