Upload folder using huggingface_hub
Browse files- .gitattributes +1 -0
- IQ3_S/Qwen3.5-35B-A3B-IQ3_S-00001-of-00002.gguf +1 -1
- IQ3_S/Qwen3.5-35B-A3B-IQ3_S-00002-of-00002.gguf +2 -2
- IQ4_XS/Qwen3.5-35B-A3B-IQ4_XS-00001-of-00002.gguf +1 -1
- IQ4_XS/Qwen3.5-35B-A3B-IQ4_XS-00002-of-00002.gguf +2 -2
- Q4_K_M/Qwen3.5-35B-A3B-Q4_K_M-00001-of-00002.gguf +1 -1
- Q4_K_M/Qwen3.5-35B-A3B-Q4_K_M-00002-of-00002.gguf +2 -2
- Q5_K_M/Qwen3.5-35B-A3B-Q5_K_M-00001-of-00002.gguf +1 -1
- Q5_K_M/Qwen3.5-35B-A3B-Q5_K_M-00002-of-00002.gguf +2 -2
- imatrix.gguf +3 -0
- kld_data/01_kld_vs_filesize.png +2 -2
- kld_data/02_ppl_vs_filesize.png +2 -2
- kld_data/aes_sedai/Qwen3.5-35B-A3B-IQ3_S.md +0 -0
- kld_data/aes_sedai/Qwen3.5-35B-A3B-IQ4_XS.md +0 -0
- kld_data/aes_sedai/Qwen3.5-35B-A3B-Q4_K_M.md +0 -0
- kld_data/aes_sedai/Qwen3.5-35B-A3B-Q5_K_M.md +0 -0
- kld_data/llm_quantization_data.csv +4 -5
.gitattributes
CHANGED
|
@@ -47,3 +47,4 @@ mmproj-Qwen3.5-35B-A3B-F32.gguf filter=lfs diff=lfs merge=lfs -text
|
|
| 47 |
mmproj-Qwen3.5-35B-A3B-Q8_0.gguf filter=lfs diff=lfs merge=lfs -text
|
| 48 |
kld_data/01_kld_vs_filesize.png filter=lfs diff=lfs merge=lfs -text
|
| 49 |
kld_data/02_ppl_vs_filesize.png filter=lfs diff=lfs merge=lfs -text
|
|
|
|
|
|
| 47 |
mmproj-Qwen3.5-35B-A3B-Q8_0.gguf filter=lfs diff=lfs merge=lfs -text
|
| 48 |
kld_data/01_kld_vs_filesize.png filter=lfs diff=lfs merge=lfs -text
|
| 49 |
kld_data/02_ppl_vs_filesize.png filter=lfs diff=lfs merge=lfs -text
|
| 50 |
+
imatrix.gguf filter=lfs diff=lfs merge=lfs -text
|
IQ3_S/Qwen3.5-35B-A3B-IQ3_S-00001-of-00002.gguf
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 10943168
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:155b733b530d9fb71ba0a1bbf7c7053593878951a16c9059f866585199c3a303
|
| 3 |
size 10943168
|
IQ3_S/Qwen3.5-35B-A3B-IQ3_S-00002-of-00002.gguf
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
-
size
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:19b3175da438bb0926972463f925a2253c0b8934e7fd597728a39fb535014fc2
|
| 3 |
+
size 13583987776
|
IQ4_XS/Qwen3.5-35B-A3B-IQ4_XS-00001-of-00002.gguf
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 10943168
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:72e58657fd2d08ac477040b2bf7230fd0503aa60aaef4bc02595e04606d2f08f
|
| 3 |
size 10943168
|
IQ4_XS/Qwen3.5-35B-A3B-IQ4_XS-00002-of-00002.gguf
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
-
size
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:83bcd0a6a08c62f6e83576008e867ce6483fdf899177d4c2d02b61bb68237c79
|
| 3 |
+
size 17609946176
|
Q4_K_M/Qwen3.5-35B-A3B-Q4_K_M-00001-of-00002.gguf
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 10943168
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:72e58657fd2d08ac477040b2bf7230fd0503aa60aaef4bc02595e04606d2f08f
|
| 3 |
size 10943168
|
Q4_K_M/Qwen3.5-35B-A3B-Q4_K_M-00002-of-00002.gguf
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
-
size
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:363f8c102fa9dccc7b486c4ea03b08f2b43888994929952601ad9f47d92b3729
|
| 3 |
+
size 22139794496
|
Q5_K_M/Qwen3.5-35B-A3B-Q5_K_M-00001-of-00002.gguf
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 10943168
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:72e58657fd2d08ac477040b2bf7230fd0503aa60aaef4bc02595e04606d2f08f
|
| 3 |
size 10943168
|
Q5_K_M/Qwen3.5-35B-A3B-Q5_K_M-00002-of-00002.gguf
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
-
size
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:cca989feaadd51be44cd73ece37f4a67c194ca861f66da59ebc137de93ef7687
|
| 3 |
+
size 26250212416
|
imatrix.gguf
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:0587dbe56de3267c2db792b115fef192baaed8565fafdfa532277a63a22f6506
|
| 3 |
+
size 108291328
|
kld_data/01_kld_vs_filesize.png
CHANGED
|
Git LFS Details
|
|
Git LFS Details
|
kld_data/02_ppl_vs_filesize.png
CHANGED
|
Git LFS Details
|
|
Git LFS Details
|
kld_data/aes_sedai/Qwen3.5-35B-A3B-IQ3_S.md
CHANGED
|
The diff for this file is too large to render.
See raw diff
|
|
|
kld_data/aes_sedai/Qwen3.5-35B-A3B-IQ4_XS.md
CHANGED
|
The diff for this file is too large to render.
See raw diff
|
|
|
kld_data/aes_sedai/Qwen3.5-35B-A3B-Q4_K_M.md
CHANGED
|
The diff for this file is too large to render.
See raw diff
|
|
|
kld_data/aes_sedai/Qwen3.5-35B-A3B-Q5_K_M.md
CHANGED
|
The diff for this file is too large to render.
See raw diff
|
|
|
kld_data/llm_quantization_data.csv
CHANGED
|
@@ -1,6 +1,5 @@
|
|
| 1 |
model_name,file_size_gb,bpw,Mean KLD_mean,0.1% KLD,0.1% Δp,1.0% KLD,1.0% Δp,10.0% KLD,10.0% Δp,25.0% Δp,5.0% KLD,5.0% Δp,75.0% Δp,90.0% KLD,90.0% Δp,95.0% KLD,95.0% Δp,99.0% KLD,99.0% Δp,99.9% KLD,99.9% Δp,"Cor(ln(PPL(Q)), ln(PPL(base)))",Device 0,Device 1,Maximum KLD,Maximum Δp,Mean KLD_std,Mean PPL(Q)-PPL(base)_mean,Mean PPL(Q)-PPL(base)_std,Mean PPL(Q)/PPL(base)_mean,Mean PPL(Q)/PPL(base)_std,Mean PPL(Q)_mean,Mean PPL(Q)_std,Mean PPL(base)_mean,Mean PPL(base)_std,Mean ln(PPL(Q)/PPL(base))_mean,Mean ln(PPL(Q)/PPL(base))_std,Mean Δp_mean,Mean Δp_std,Median KLD,Median Δp,Minimum KLD,Minimum Δp,RMS Δp_mean,RMS Δp_std,Same top p_mean,Same top p_std,file_path,file_size_gib,ggml_cuda_init,kl_divergence,llama_context,llama_kv_cache,llama_memory_breakdown_print,llama_memory_recurrent,llama_model_loader,llama_params_fit,llama_params_fit_impl,llama_perf_context_print,load,load_tensors,print_info,sched_reserve,system_info
|
| 2 |
-
Qwen3.5-35B-A3B-IQ3_S (aes_sedai),13.
|
| 3 |
-
Qwen3.5-35B-A3B-IQ4_XS (aes_sedai),17.
|
| 4 |
-
Qwen3.5-35B-A3B-Q4_K_M (aes_sedai),22.
|
| 5 |
-
Qwen3.5-35B-A3B-Q5_K_M (aes_sedai),26.
|
| 6 |
-
Qwen3.5-35B-A3B-Q8_0 (aes_sedai),36.893769072640005,8.52,0.004627,-6e-06,-14.494,2e-06,-5.632,7.1e-05,-1.463,-0.336,2e-05,-2.472,0.309,0.008926,1.41,0.014622,2.415,0.042379,5.458,0.155776,14.119,99.9,30908.6,30908.6,12.434992,80.935,0.000119,0.001061,0.001904,1.000162,0.000291,6.535132,0.041571,6.534071,0.04154,0.000162,0.000291,-0.013,0.005,0.001935,-0.0,-7e-05,-91.893,1.978,0.03,96.794,0.046,kld/Qwen3.5-35B-A3B/wiki_test_raw/aes_sedai/Qwen3.5-35B-A3B-Q8_0.md,34.36,2.0,58051240968.0,7.58,132.0,-587515072.0,1184.25,-80432.0,5.32,-1.3090161174816154e+16,71.0,1.7581,113320.84,256.0,47.101,4.848568601128313e+47
|
|
|
|
| 1 |
model_name,file_size_gb,bpw,Mean KLD_mean,0.1% KLD,0.1% Δp,1.0% KLD,1.0% Δp,10.0% KLD,10.0% Δp,25.0% Δp,5.0% KLD,5.0% Δp,75.0% Δp,90.0% KLD,90.0% Δp,95.0% KLD,95.0% Δp,99.0% KLD,99.0% Δp,99.9% KLD,99.9% Δp,"Cor(ln(PPL(Q)), ln(PPL(base)))",Device 0,Device 1,Maximum KLD,Maximum Δp,Mean KLD_std,Mean PPL(Q)-PPL(base)_mean,Mean PPL(Q)-PPL(base)_std,Mean PPL(Q)/PPL(base)_mean,Mean PPL(Q)/PPL(base)_std,Mean PPL(Q)_mean,Mean PPL(Q)_std,Mean PPL(base)_mean,Mean PPL(base)_std,Mean ln(PPL(Q)/PPL(base))_mean,Mean ln(PPL(Q)/PPL(base))_std,Mean Δp_mean,Mean Δp_std,Median KLD,Median Δp,Minimum KLD,Minimum Δp,RMS Δp_mean,RMS Δp_std,Same top p_mean,Same top p_std,file_path,file_size_gib,ggml_cuda_init,kl_divergence,llama_context,llama_kv_cache,llama_memory_breakdown_print,llama_memory_recurrent,llama_model_loader,llama_params_fit,llama_params_fit_impl,llama_perf_context_print,load,load_tensors,print_info,sched_reserve,system_info
|
| 2 |
+
Qwen3.5-35B-A3B-IQ3_S (aes_sedai),13.5828340736,3.14,0.061926,9e-06,-67.902,5.5e-05,-30.275,0.001044,-6.728,-1.878,0.000321,-11.669,0.628,0.129668,3.954,0.220491,7.105,0.634623,16.226,1.955343,37.068,98.68,,,10.490875,98.158,0.000402,0.385178,0.007646,1.058937,0.001107,6.920534,0.044619,6.535357,0.041544,0.057266,0.001045,-1.099,0.019,0.02653,-0.056,-1.7e-05,-98.703,7.567,0.054,89.355,0.08,kld/Qwen3.5-35B-A3B/wiki-test-raw/aes_sedai/Qwen3.5-35B-A3B-IQ3_S.md,12.65,2.0,580512819216.0,15.16,180.0,-5913970193.0,1469.0,-240.0,0.51,24022.192373,35.0,1.7581,16171.97,256.0,118.904,4.848561200112841e+50
|
| 3 |
+
Qwen3.5-35B-A3B-IQ4_XS (aes_sedai),17.609365913599998,4.06,0.02413,2e-06,-41.876,1.8e-05,-15.476,0.000377,-3.688,-0.943,0.000114,-6.237,0.519,0.049274,2.693,0.083029,4.828,0.245648,11.398,0.815935,28.995,99.49,,,18.38467,96.562,0.00025,0.097635,0.004284,1.01494,0.000651,6.632992,0.042286,6.535357,0.041544,0.014829,0.000641,-0.358,0.012,0.009995,-0.009,-3.3e-05,-85.302,4.575,0.04,93.285,0.065,kld/Qwen3.5-35B-A3B/wiki-test-raw/aes_sedai/Qwen3.5-35B-A3B-IQ4_XS.md,16.4,2.0,580512819216.0,15.16,180.0,-7085150193.0,1469.0,-440.0,0.55,27744.192373,35.0,1.7581,18001.22,256.0,105.594,4.848561200112841e+50
|
| 4 |
+
Qwen3.5-35B-A3B-Q4_K_M (aes_sedai),22.140556410880002,5.11,0.010121,-2e-06,-22.936,6e-06,-8.951,0.00016,-2.185,-0.518,4.7e-05,-3.751,0.425,0.020696,1.975,0.034738,3.434,0.098049,8.135,0.322683,20.616,99.79,,,11.692275,96.85,0.000122,0.028815,0.002718,1.004409,0.000415,6.564172,0.041821,6.535357,0.041544,0.004399,0.000413,-0.076,0.008,0.004253,-0.0,-0.000146,-62.979,2.921,0.03,95.563,0.054,kld/Qwen3.5-35B-A3B/wiki-test-raw/aes_sedai/Qwen3.5-35B-A3B-Q4_K_M.md,20.62,2.0,580512819216.0,15.16,180.0,-7085150193.0,1469.0,-540.0,0.53,32064.192373,35.0,1.7581,110053.22,256.0,132.584,4.848561200112841e+50
|
| 5 |
+
Qwen3.5-35B-A3B-Q5_K_M (aes_sedai),26.2529875968,6.06,0.006174,-5e-06,-16.816,3e-06,-6.642,0.000101,-1.706,-0.401,2.9e-05,-2.858,0.355,0.012446,1.614,0.020427,2.788,0.05914,6.473,0.218745,16.348,99.87,,,13.798801,94.344,0.000108,-0.001192,0.002143,0.999818,0.000328,6.534165,0.041552,6.535357,0.041544,-0.000182,0.000328,-0.024,0.006,0.002631,-0.0,-7.3e-05,-73.266,2.288,0.028,96.374,0.049,kld/Qwen3.5-35B-A3B/wiki-test-raw/aes_sedai/Qwen3.5-35B-A3B-Q5_K_M.md,24.45,2.0,580512819216.0,15.16,180.0,-7085150193.0,1469.0,-640.0,1.45,35984.192373,35.0,1.7581,111915.22,256.0,94.364,4.848561200112841e+50
|
|
|