forked from ggerganov/llama.cpp
-
Notifications
You must be signed in to change notification settings - Fork 1
Commit
This commit does not belong to any branch on this repository, and may belong to a fork outside of the repository.
IQ1_M: 1.75 bpw quantization (ggerganov#6302)
* iq1_m: basics * iq1_m: basics-2 * iq1_m: CUDA dequantize works Very 1st shot I get PPL = 9.76 for LLaMA-v2-7B. * iq1_m: separate shifts for each group of 8 in a block We get PPL(LLaMA-v2-7B ) = 9.2810 PPL(LLaMA-v2-13B) = 6.8105 Not bad, but slightly higher than sqrt(PPL(IQ1_S) * PPL(IQ2_XXS)) which is the expected outcome given that IQ1_M is halfway between IQ1_S and IQ2_XXS in terms of bpw. From this, we would expect PPL = 9.14 for LLaMA-v2-7B PPL = 6.63 for LLaMA-v2-13B * iq1_m: go to 3-bit scales There is slight increase in PPL, but the 0.0625 bpw reduction in size is totally worth it. We now have PPL(LLaMA-v2-7B ) = 9.4469 at 1.96 bpw PPL(LLaMA-v2-13B) = 6.8717 at 1.93 bpw PPL(LLaMA-v2-70B) = 4.8568 at 1.85 bpw * iq1_m: scalar dot product * iq1_m: AVX2 dot product * iq1_m: very slightly faster AVX2 dot product * iq1_m: ARM_NEON dot product Works, but very slow (10.5 t/s) * iq1_m: Metal - dequantize works, dot product does not * iq1_m: Metal now works About the same performance as iq1_s. * iq1_m: minor * iq1_m: checking pure iq1_m quantization It is pretty bad: PPL(LLaMA-v2-7B) = 34 if we quantize output.weight with Q4_K. * iiq1_m: slightly faster ARM_NEON dot product 10.5 t/s -> 11.65 t/s * iq1_m: faster ARM_NEON dot product 11.65 t/s -> 14.9 t/s * iq1_m: another minor ARM_NEON dot product improvement 14.9 -> 15.0 t/s * iq1_m: small PPL improvement via super-block scale adjustment After quantizing block scales redo the super-block scale fit. PPL(LLaMA-v2-7B ) = 9.3346 PPL(LLaMA-v2-13B) = 6.8419 PPL(LLaMA-v2-70B) = 4.8294 PPL(Mistral-7B ) = 8.1624 * iq1_m: adapt to CUDA refactoring * iq1_m: remove unused variable We have progressed to warnings being errors. * iq1_m: add to backend-ops tests * iq1_m: fix Windows ARM * iq1_m: use common definition of iq1m_scale_t * cuda: assert -> NO_DEVICE_CODE * iq1_M: PR comments --------- Co-authored-by: Iwan Kawrakow <iwan.kawrakow@gmail.com>
- Loading branch information
Showing
16 changed files
with
1,006 additions
and
125 deletions.
There are no files selected for viewing
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Oops, something went wrong.