Hugging Face's logo

M5Stack
/
Qwen3-1.7B-GPTQ-Int4-axmodel

Model card Files
xet
 Community
LittleMouse commited on
Commit
c9557ca
·
1 Parent(s): 11535d8

Upload shell

Browse files
Files changed (32) hide show
  1. {Qwen3-1.7B-GPTQ-Int8-ax8850 → Qwen3-1.7B-GPTQ-Int4-ax8850}/model.embed_tokens.weight.bfloat16.bin +0 -0
  2. {Qwen3-1.7B-GPTQ-Int8-ax8850 → Qwen3-1.7B-GPTQ-Int4-ax8850}/qwen3_p128_l0_together.axmodel +0 -0
  3. {Qwen3-1.7B-GPTQ-Int8-ax8850 → Qwen3-1.7B-GPTQ-Int4-ax8850}/qwen3_p128_l10_together.axmodel +0 -0
  4. {Qwen3-1.7B-GPTQ-Int8-ax8850 → Qwen3-1.7B-GPTQ-Int4-ax8850}/qwen3_p128_l11_together.axmodel +0 -0
  5. {Qwen3-1.7B-GPTQ-Int8-ax8850 → Qwen3-1.7B-GPTQ-Int4-ax8850}/qwen3_p128_l12_together.axmodel +0 -0
  6. {Qwen3-1.7B-GPTQ-Int8-ax8850 → Qwen3-1.7B-GPTQ-Int4-ax8850}/qwen3_p128_l13_together.axmodel +0 -0
  7. {Qwen3-1.7B-GPTQ-Int8-ax8850 → Qwen3-1.7B-GPTQ-Int4-ax8850}/qwen3_p128_l14_together.axmodel +0 -0
  8. {Qwen3-1.7B-GPTQ-Int8-ax8850 → Qwen3-1.7B-GPTQ-Int4-ax8850}/qwen3_p128_l15_together.axmodel +0 -0
  9. {Qwen3-1.7B-GPTQ-Int8-ax8850 → Qwen3-1.7B-GPTQ-Int4-ax8850}/qwen3_p128_l16_together.axmodel +0 -0
  10. {Qwen3-1.7B-GPTQ-Int8-ax8850 → Qwen3-1.7B-GPTQ-Int4-ax8850}/qwen3_p128_l17_together.axmodel +0 -0
  11. {Qwen3-1.7B-GPTQ-Int8-ax8850 → Qwen3-1.7B-GPTQ-Int4-ax8850}/qwen3_p128_l18_together.axmodel +0 -0
  12. {Qwen3-1.7B-GPTQ-Int8-ax8850 → Qwen3-1.7B-GPTQ-Int4-ax8850}/qwen3_p128_l19_together.axmodel +0 -0
  13. {Qwen3-1.7B-GPTQ-Int8-ax8850 → Qwen3-1.7B-GPTQ-Int4-ax8850}/qwen3_p128_l1_together.axmodel +0 -0
  14. {Qwen3-1.7B-GPTQ-Int8-ax8850 → Qwen3-1.7B-GPTQ-Int4-ax8850}/qwen3_p128_l20_together.axmodel +0 -0
  15. {Qwen3-1.7B-GPTQ-Int8-ax8850 → Qwen3-1.7B-GPTQ-Int4-ax8850}/qwen3_p128_l21_together.axmodel +0 -0
  16. {Qwen3-1.7B-GPTQ-Int8-ax8850 → Qwen3-1.7B-GPTQ-Int4-ax8850}/qwen3_p128_l22_together.axmodel +0 -0
  17. {Qwen3-1.7B-GPTQ-Int8-ax8850 → Qwen3-1.7B-GPTQ-Int4-ax8850}/qwen3_p128_l23_together.axmodel +0 -0
  18. {Qwen3-1.7B-GPTQ-Int8-ax8850 → Qwen3-1.7B-GPTQ-Int4-ax8850}/qwen3_p128_l24_together.axmodel +0 -0
  19. {Qwen3-1.7B-GPTQ-Int8-ax8850 → Qwen3-1.7B-GPTQ-Int4-ax8850}/qwen3_p128_l25_together.axmodel +0 -0
  20. {Qwen3-1.7B-GPTQ-Int8-ax8850 → Qwen3-1.7B-GPTQ-Int4-ax8850}/qwen3_p128_l26_together.axmodel +0 -0
  21. {Qwen3-1.7B-GPTQ-Int8-ax8850 → Qwen3-1.7B-GPTQ-Int4-ax8850}/qwen3_p128_l27_together.axmodel +0 -0
  22. {Qwen3-1.7B-GPTQ-Int8-ax8850 → Qwen3-1.7B-GPTQ-Int4-ax8850}/qwen3_p128_l2_together.axmodel +0 -0
  23. {Qwen3-1.7B-GPTQ-Int8-ax8850 → Qwen3-1.7B-GPTQ-Int4-ax8850}/qwen3_p128_l3_together.axmodel +0 -0
  24. {Qwen3-1.7B-GPTQ-Int8-ax8850 → Qwen3-1.7B-GPTQ-Int4-ax8850}/qwen3_p128_l4_together.axmodel +0 -0
  25. {Qwen3-1.7B-GPTQ-Int8-ax8850 → Qwen3-1.7B-GPTQ-Int4-ax8850}/qwen3_p128_l5_together.axmodel +0 -0
  26. {Qwen3-1.7B-GPTQ-Int8-ax8850 → Qwen3-1.7B-GPTQ-Int4-ax8850}/qwen3_p128_l6_together.axmodel +0 -0
  27. {Qwen3-1.7B-GPTQ-Int8-ax8850 → Qwen3-1.7B-GPTQ-Int4-ax8850}/qwen3_p128_l7_together.axmodel +0 -0
  28. {Qwen3-1.7B-GPTQ-Int8-ax8850 → Qwen3-1.7B-GPTQ-Int4-ax8850}/qwen3_p128_l8_together.axmodel +0 -0
  29. {Qwen3-1.7B-GPTQ-Int8-ax8850 → Qwen3-1.7B-GPTQ-Int4-ax8850}/qwen3_p128_l9_together.axmodel +0 -0
  30. {Qwen3-1.7B-GPTQ-Int8-ax8850 → Qwen3-1.7B-GPTQ-Int4-ax8850}/qwen3_post.axmodel +0 -0
  31. run_qwen3_1.7b_int8_ctx_axcl_aarch64.sh +12 -0
  32. run_qwen3_1.7b_int8_ctx_axcl_aarch_api.sh +11 -0
{Qwen3-1.7B-GPTQ-Int8-ax8850 → Qwen3-1.7B-GPTQ-Int4-ax8850}/model.embed_tokens.weight.bfloat16.bin RENAMED
File without changes
{Qwen3-1.7B-GPTQ-Int8-ax8850 → Qwen3-1.7B-GPTQ-Int4-ax8850}/qwen3_p128_l0_together.axmodel RENAMED
File without changes
{Qwen3-1.7B-GPTQ-Int8-ax8850 → Qwen3-1.7B-GPTQ-Int4-ax8850}/qwen3_p128_l10_together.axmodel RENAMED
File without changes
{Qwen3-1.7B-GPTQ-Int8-ax8850 → Qwen3-1.7B-GPTQ-Int4-ax8850}/qwen3_p128_l11_together.axmodel RENAMED
File without changes
{Qwen3-1.7B-GPTQ-Int8-ax8850 → Qwen3-1.7B-GPTQ-Int4-ax8850}/qwen3_p128_l12_together.axmodel RENAMED
File without changes
{Qwen3-1.7B-GPTQ-Int8-ax8850 → Qwen3-1.7B-GPTQ-Int4-ax8850}/qwen3_p128_l13_together.axmodel RENAMED
File without changes
{Qwen3-1.7B-GPTQ-Int8-ax8850 → Qwen3-1.7B-GPTQ-Int4-ax8850}/qwen3_p128_l14_together.axmodel RENAMED
File without changes
{Qwen3-1.7B-GPTQ-Int8-ax8850 → Qwen3-1.7B-GPTQ-Int4-ax8850}/qwen3_p128_l15_together.axmodel RENAMED
File without changes
{Qwen3-1.7B-GPTQ-Int8-ax8850 → Qwen3-1.7B-GPTQ-Int4-ax8850}/qwen3_p128_l16_together.axmodel RENAMED
File without changes
{Qwen3-1.7B-GPTQ-Int8-ax8850 → Qwen3-1.7B-GPTQ-Int4-ax8850}/qwen3_p128_l17_together.axmodel RENAMED
File without changes
{Qwen3-1.7B-GPTQ-Int8-ax8850 → Qwen3-1.7B-GPTQ-Int4-ax8850}/qwen3_p128_l18_together.axmodel RENAMED
File without changes
{Qwen3-1.7B-GPTQ-Int8-ax8850 → Qwen3-1.7B-GPTQ-Int4-ax8850}/qwen3_p128_l19_together.axmodel RENAMED
File without changes
{Qwen3-1.7B-GPTQ-Int8-ax8850 → Qwen3-1.7B-GPTQ-Int4-ax8850}/qwen3_p128_l1_together.axmodel RENAMED
File without changes
{Qwen3-1.7B-GPTQ-Int8-ax8850 → Qwen3-1.7B-GPTQ-Int4-ax8850}/qwen3_p128_l20_together.axmodel RENAMED
File without changes
{Qwen3-1.7B-GPTQ-Int8-ax8850 → Qwen3-1.7B-GPTQ-Int4-ax8850}/qwen3_p128_l21_together.axmodel RENAMED
File without changes
{Qwen3-1.7B-GPTQ-Int8-ax8850 → Qwen3-1.7B-GPTQ-Int4-ax8850}/qwen3_p128_l22_together.axmodel RENAMED
File without changes
{Qwen3-1.7B-GPTQ-Int8-ax8850 → Qwen3-1.7B-GPTQ-Int4-ax8850}/qwen3_p128_l23_together.axmodel RENAMED
File without changes
{Qwen3-1.7B-GPTQ-Int8-ax8850 → Qwen3-1.7B-GPTQ-Int4-ax8850}/qwen3_p128_l24_together.axmodel RENAMED
File without changes
{Qwen3-1.7B-GPTQ-Int8-ax8850 → Qwen3-1.7B-GPTQ-Int4-ax8850}/qwen3_p128_l25_together.axmodel RENAMED
File without changes
{Qwen3-1.7B-GPTQ-Int8-ax8850 → Qwen3-1.7B-GPTQ-Int4-ax8850}/qwen3_p128_l26_together.axmodel RENAMED
File without changes
{Qwen3-1.7B-GPTQ-Int8-ax8850 → Qwen3-1.7B-GPTQ-Int4-ax8850}/qwen3_p128_l27_together.axmodel RENAMED
File without changes
{Qwen3-1.7B-GPTQ-Int8-ax8850 → Qwen3-1.7B-GPTQ-Int4-ax8850}/qwen3_p128_l2_together.axmodel RENAMED
File without changes
{Qwen3-1.7B-GPTQ-Int8-ax8850 → Qwen3-1.7B-GPTQ-Int4-ax8850}/qwen3_p128_l3_together.axmodel RENAMED
File without changes
{Qwen3-1.7B-GPTQ-Int8-ax8850 → Qwen3-1.7B-GPTQ-Int4-ax8850}/qwen3_p128_l4_together.axmodel RENAMED
File without changes
{Qwen3-1.7B-GPTQ-Int8-ax8850 → Qwen3-1.7B-GPTQ-Int4-ax8850}/qwen3_p128_l5_together.axmodel RENAMED
File without changes
{Qwen3-1.7B-GPTQ-Int8-ax8850 → Qwen3-1.7B-GPTQ-Int4-ax8850}/qwen3_p128_l6_together.axmodel RENAMED
File without changes
{Qwen3-1.7B-GPTQ-Int8-ax8850 → Qwen3-1.7B-GPTQ-Int4-ax8850}/qwen3_p128_l7_together.axmodel RENAMED
File without changes
{Qwen3-1.7B-GPTQ-Int8-ax8850 → Qwen3-1.7B-GPTQ-Int4-ax8850}/qwen3_p128_l8_together.axmodel RENAMED
File without changes
{Qwen3-1.7B-GPTQ-Int8-ax8850 → Qwen3-1.7B-GPTQ-Int4-ax8850}/qwen3_p128_l9_together.axmodel RENAMED
File without changes
{Qwen3-1.7B-GPTQ-Int8-ax8850 → Qwen3-1.7B-GPTQ-Int4-ax8850}/qwen3_post.axmodel RENAMED
File without changes
run_qwen3_1.7b_int8_ctx_axcl_aarch64.sh ADDED
@@ -0,0 +1,12 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ ./main_axcl_aarch64 \
2
+ --system_prompt "You are Qwen, created by Alibaba Cloud. You are a helpful assistant." \
3
+ --template_filename_axmodel "Qwen3-1.7B-GPTQ-Int4-ax8850/qwen3_p128_l%d_together.axmodel" \
4
+ --axmodel_num 28 \
5
+ --url_tokenizer_model "http://127.0.0.1:12345" \
6
+ --filename_post_axmodel Qwen3-1.7B-GPTQ-Int4-ax8850/qwen3_post.axmodel \
7
+ --filename_tokens_embed Qwen3-1.7B-GPTQ-Int4-ax8850/model.embed_tokens.weight.bfloat16.bin \
8
+ --tokens_embed_num 151936 \
9
+ --tokens_embed_size 2048 \
10
+ --use_mmap_load_embed 1 \
11
+ --live_print 1 \
12
+ --devices 0
run_qwen3_1.7b_int8_ctx_axcl_aarch_api.sh ADDED
@@ -0,0 +1,11 @@
 
 
 
 
 
 
 
 
 
 
 
 
1
+ ./main_api_axcl_aarch64 \
2
+ --system_prompt "You are Qwen, created by Alibaba Cloud. You are a helpful assistant." \
3
+ --template_filename_axmodel "Qwen3-1.7B-GPTQ-Int4-ax8850/qwen3_p128_l%d_together.axmodel" \
4
+ --axmodel_num 28 \
5
+ --url_tokenizer_model "http://127.0.0.1:12345" \
6
+ --filename_post_axmodel Qwen3-1.7B-GPTQ-Int4-ax8850/qwen3_post.axmodel \
7
+ --filename_tokens_embed Qwen3-1.7B-GPTQ-Int4-ax8850/model.embed_tokens.weight.bfloat16.bin \
8
+ --tokens_embed_num 151936 \
9
+ --tokens_embed_size 2048 \
10
+ --use_mmap_load_embed 1 \
11
+ --devices 0