Hugging Face's logo

imi2
/
QNN-HTP-LLM-Genie-models

Model card Files Community
1
imi2 commited on
Commit
d9e17ee
·
verified ·
1 Parent(s): 36b09f9

Upload 71 files

Browse files
This view is limited to 50 files because it contains too many changes.   See raw diff
Files changed (50) hide show
  1. .gitattributes +20 -0
  2. genie_bundle_llama_v3_2_3b_gen_2/genie-t2t-run +0 -0
  3. genie_bundle_llama_v3_2_3b_gen_2/genie_config.json +67 -0
  4. genie_bundle_llama_v3_2_3b_gen_2/htp_backend_ext_config.json +21 -0
  5. genie_bundle_llama_v3_2_3b_gen_2/libCalculator_skel.so +0 -0
  6. genie_bundle_llama_v3_2_3b_gen_2/libGenie.so +3 -0
  7. genie_bundle_llama_v3_2_3b_gen_2/libPlatformValidatorShared.so +3 -0
  8. genie_bundle_llama_v3_2_3b_gen_2/libQnnChrometraceProfilingReader.so +0 -0
  9. genie_bundle_llama_v3_2_3b_gen_2/libQnnCpu.so +3 -0
  10. genie_bundle_llama_v3_2_3b_gen_2/libQnnDsp.so +3 -0
  11. genie_bundle_llama_v3_2_3b_gen_2/libQnnDspNetRunExtensions.so +0 -0
  12. genie_bundle_llama_v3_2_3b_gen_2/libQnnDspV66CalculatorStub.so +0 -0
  13. genie_bundle_llama_v3_2_3b_gen_2/libQnnDspV66Stub.so +0 -0
  14. genie_bundle_llama_v3_2_3b_gen_2/libQnnGenAiTransformer.so +0 -0
  15. genie_bundle_llama_v3_2_3b_gen_2/libQnnGenAiTransformerCpuOpPkg.so +0 -0
  16. genie_bundle_llama_v3_2_3b_gen_2/libQnnGenAiTransformerModel.so +0 -0
  17. genie_bundle_llama_v3_2_3b_gen_2/libQnnGpu.so +3 -0
  18. genie_bundle_llama_v3_2_3b_gen_2/libQnnGpuNetRunExtensions.so +0 -0
  19. genie_bundle_llama_v3_2_3b_gen_2/libQnnGpuProfilingReader.so +0 -0
  20. genie_bundle_llama_v3_2_3b_gen_2/libQnnHta.so +3 -0
  21. genie_bundle_llama_v3_2_3b_gen_2/libQnnHtaNetRunExtensions.so +0 -0
  22. genie_bundle_llama_v3_2_3b_gen_2/libQnnHtp.so +3 -0
  23. genie_bundle_llama_v3_2_3b_gen_2/libQnnHtpNetRunExtensions.so +0 -0
  24. genie_bundle_llama_v3_2_3b_gen_2/libQnnHtpOptraceProfilingReader.so +3 -0
  25. genie_bundle_llama_v3_2_3b_gen_2/libQnnHtpPrepare.so +3 -0
  26. genie_bundle_llama_v3_2_3b_gen_2/libQnnHtpProfilingReader.so +0 -0
  27. genie_bundle_llama_v3_2_3b_gen_2/libQnnHtpV68CalculatorStub.so +0 -0
  28. genie_bundle_llama_v3_2_3b_gen_2/libQnnHtpV68Stub.so +0 -0
  29. genie_bundle_llama_v3_2_3b_gen_2/libQnnHtpV69CalculatorStub.so +0 -0
  30. genie_bundle_llama_v3_2_3b_gen_2/libQnnHtpV69Stub.so +0 -0
  31. genie_bundle_llama_v3_2_3b_gen_2/libQnnHtpV73.so +3 -0
  32. genie_bundle_llama_v3_2_3b_gen_2/libQnnHtpV73CalculatorStub.so +0 -0
  33. genie_bundle_llama_v3_2_3b_gen_2/libQnnHtpV73QemuDriver.so +3 -0
  34. genie_bundle_llama_v3_2_3b_gen_2/libQnnHtpV73Skel.so +3 -0
  35. genie_bundle_llama_v3_2_3b_gen_2/libQnnHtpV73Stub.so +0 -0
  36. genie_bundle_llama_v3_2_3b_gen_2/libQnnHtpV75CalculatorStub.so +0 -0
  37. genie_bundle_llama_v3_2_3b_gen_2/libQnnHtpV75Stub.so +0 -0
  38. genie_bundle_llama_v3_2_3b_gen_2/libQnnHtpV79CalculatorStub.so +0 -0
  39. genie_bundle_llama_v3_2_3b_gen_2/libQnnHtpV79Stub.so +0 -0
  40. genie_bundle_llama_v3_2_3b_gen_2/libQnnJsonProfilingReader.so +0 -0
  41. genie_bundle_llama_v3_2_3b_gen_2/libQnnLpai.so +0 -0
  42. genie_bundle_llama_v3_2_3b_gen_2/libQnnLpaiNetRunExtensions.so +0 -0
  43. genie_bundle_llama_v3_2_3b_gen_2/libQnnLpaiV79Stub.so +0 -0
  44. genie_bundle_llama_v3_2_3b_gen_2/libQnnModelDlc.so +3 -0
  45. genie_bundle_llama_v3_2_3b_gen_2/libQnnNetRunDirectV79Stub.so +0 -0
  46. genie_bundle_llama_v3_2_3b_gen_2/libQnnSaver.so +0 -0
  47. genie_bundle_llama_v3_2_3b_gen_2/libQnnSystem.so +0 -0
  48. genie_bundle_llama_v3_2_3b_gen_2/libQnnTFLiteDelegate.so +0 -0
  49. genie_bundle_llama_v3_2_3b_gen_2/libSNPE.so +3 -0
  50. genie_bundle_llama_v3_2_3b_gen_2/libSnpeDspV66Stub.so +3 -0
.gitattributes CHANGED
@@ -53,3 +53,23 @@ genie_bundle_llama_v3_1_8b_Supernova_Lite_4096_gen_2/libSnpeDspV66Stub.so filter
53
  genie_bundle_llama_v3_1_8b_Supernova_Lite_4096_gen_2/libSnpeHta.so filter=lfs diff=lfs merge=lfs -text
54
  genie_bundle_llama_v3_1_8b_Supernova_Lite_4096_gen_2/libSnpeHtpPrepare.so filter=lfs diff=lfs merge=lfs -text
55
  genie_bundle_llama_v3_1_8b_Supernova_Lite_4096_gen_2/libSnpeHtpV73Skel.so filter=lfs diff=lfs merge=lfs -text
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
53
  genie_bundle_llama_v3_1_8b_Supernova_Lite_4096_gen_2/libSnpeHta.so filter=lfs diff=lfs merge=lfs -text
54
  genie_bundle_llama_v3_1_8b_Supernova_Lite_4096_gen_2/libSnpeHtpPrepare.so filter=lfs diff=lfs merge=lfs -text
55
  genie_bundle_llama_v3_1_8b_Supernova_Lite_4096_gen_2/libSnpeHtpV73Skel.so filter=lfs diff=lfs merge=lfs -text
56
+ genie_bundle_llama_v3_2_3b_gen_2/libGenie.so filter=lfs diff=lfs merge=lfs -text
57
+ genie_bundle_llama_v3_2_3b_gen_2/libhta_hexagon_runtime_qnn.so filter=lfs diff=lfs merge=lfs -text
58
+ genie_bundle_llama_v3_2_3b_gen_2/libhta_hexagon_runtime_snpe.so filter=lfs diff=lfs merge=lfs -text
59
+ genie_bundle_llama_v3_2_3b_gen_2/libPlatformValidatorShared.so filter=lfs diff=lfs merge=lfs -text
60
+ genie_bundle_llama_v3_2_3b_gen_2/libQnnCpu.so filter=lfs diff=lfs merge=lfs -text
61
+ genie_bundle_llama_v3_2_3b_gen_2/libQnnDsp.so filter=lfs diff=lfs merge=lfs -text
62
+ genie_bundle_llama_v3_2_3b_gen_2/libQnnGpu.so filter=lfs diff=lfs merge=lfs -text
63
+ genie_bundle_llama_v3_2_3b_gen_2/libQnnHta.so filter=lfs diff=lfs merge=lfs -text
64
+ genie_bundle_llama_v3_2_3b_gen_2/libQnnHtp.so filter=lfs diff=lfs merge=lfs -text
65
+ genie_bundle_llama_v3_2_3b_gen_2/libQnnHtpOptraceProfilingReader.so filter=lfs diff=lfs merge=lfs -text
66
+ genie_bundle_llama_v3_2_3b_gen_2/libQnnHtpPrepare.so filter=lfs diff=lfs merge=lfs -text
67
+ genie_bundle_llama_v3_2_3b_gen_2/libQnnHtpV73.so filter=lfs diff=lfs merge=lfs -text
68
+ genie_bundle_llama_v3_2_3b_gen_2/libQnnHtpV73QemuDriver.so filter=lfs diff=lfs merge=lfs -text
69
+ genie_bundle_llama_v3_2_3b_gen_2/libQnnHtpV73Skel.so filter=lfs diff=lfs merge=lfs -text
70
+ genie_bundle_llama_v3_2_3b_gen_2/libQnnModelDlc.so filter=lfs diff=lfs merge=lfs -text
71
+ genie_bundle_llama_v3_2_3b_gen_2/libSNPE.so filter=lfs diff=lfs merge=lfs -text
72
+ genie_bundle_llama_v3_2_3b_gen_2/libSnpeDspV66Stub.so filter=lfs diff=lfs merge=lfs -text
73
+ genie_bundle_llama_v3_2_3b_gen_2/libSnpeHta.so filter=lfs diff=lfs merge=lfs -text
74
+ genie_bundle_llama_v3_2_3b_gen_2/libSnpeHtpPrepare.so filter=lfs diff=lfs merge=lfs -text
75
+ genie_bundle_llama_v3_2_3b_gen_2/libSnpeHtpV73Skel.so filter=lfs diff=lfs merge=lfs -text
genie_bundle_llama_v3_2_3b_gen_2/genie-t2t-run ADDED
Binary file (504 kB). View file
 
genie_bundle_llama_v3_2_3b_gen_2/genie_config.json ADDED
@@ -0,0 +1,67 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "dialog": {
3
+ "version": 1,
4
+ "type": "basic",
5
+ "context": {
6
+ "version": 1,
7
+ "size": 2048,
8
+ "n-vocab": 128256,
9
+ "bos-token": -1,
10
+ "eos-token": [128001, 128009, 128008]
11
+ },
12
+ "sampler": {
13
+ "version": 1,
14
+ "seed": 42,
15
+ "temp": 0.8,
16
+ "top-k": 40,
17
+ "top-p": 0.95
18
+ },
19
+ "tokenizer": {
20
+ "version": 1,
21
+ "path": "tokenizer.json"
22
+ },
23
+ "engine": {
24
+ "version": 1,
25
+ "n-threads": 3,
26
+ "backend": {
27
+ "version": 1,
28
+ "type": "QnnHtp",
29
+ "QnnHtp": {
30
+ "version": 1,
31
+ "use-mmap": true,
32
+ "spill-fill-bufsize": 0,
33
+ "mmap-budget": 0,
34
+ "poll": true,
35
+ "cpu-mask": "0xe0",
36
+ "kv-dim": 128,
37
+ "allow-async-init": false
38
+ },
39
+ "extensions": "htp_backend_ext_config.json"
40
+ },
41
+ "model": {
42
+ "version": 1,
43
+ "type": "binary",
44
+ "binary": {
45
+ "version": 1,
46
+ "ctx-bins": [
47
+ "llama_v3_2_3b_chat_quantized_part_1_of_3.bin",
48
+ "llama_v3_2_3b_chat_quantized_part_2_of_3.bin",
49
+ "llama_v3_2_3b_chat_quantized_part_3_of_3.bin"
50
+ ]
51
+ },
52
+ "positional-encoding": {
53
+ "type": "rope",
54
+ "rope-dim": 64,
55
+ "rope-theta": 500000,
56
+ "rope-scaling": {
57
+ "rope-type": "llama3",
58
+ "factor": 8.0,
59
+ "low-freq-factor": 1.0,
60
+ "high-freq-factor": 4.0,
61
+ "original-max-position-embeddings": 8192
62
+ }
63
+ }
64
+ }
65
+ }
66
+ }
67
+ }
genie_bundle_llama_v3_2_3b_gen_2/htp_backend_ext_config.json ADDED
@@ -0,0 +1,21 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "devices": [
3
+ {
4
+ "soc_model": 43,
5
+ "dsp_arch": "v73",
6
+ "cores": [
7
+ {
8
+ "core_id": 0,
9
+ "perf_profile": "burst",
10
+ "rpc_control_latency": 100
11
+ }
12
+ ]
13
+ }
14
+ ],
15
+ "memory": {
16
+ "mem_type": "shared_buffer"
17
+ },
18
+ "context": {
19
+ "weight_sharing_enabled": true
20
+ }
21
+ }
genie_bundle_llama_v3_2_3b_gen_2/libCalculator_skel.so ADDED
Binary file (7.12 kB). View file
 
genie_bundle_llama_v3_2_3b_gen_2/libGenie.so ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:a0ebb0d3f882df0b10497392b0e23339e6bf8b823a9988dc1913882b83148b19
3
+ size 5914016
genie_bundle_llama_v3_2_3b_gen_2/libPlatformValidatorShared.so ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:03fd62b627cd67318c4a6663cc75e74d2a43e47fef0c419b579dca414c4d2d42
3
+ size 12046864
genie_bundle_llama_v3_2_3b_gen_2/libQnnChrometraceProfilingReader.so ADDED
Binary file (590 kB). View file
 
genie_bundle_llama_v3_2_3b_gen_2/libQnnCpu.so ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:4d19429336110dfd3d4c2c2b684eb937891d41d22cdf05cf430bdfaeccd7b5f4
3
+ size 5559184
genie_bundle_llama_v3_2_3b_gen_2/libQnnDsp.so ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:20444eb13d4d1430882a52ff664fe5e9665aeac85f4e59863f15bb25c056dec6
3
+ size 1397264
genie_bundle_llama_v3_2_3b_gen_2/libQnnDspNetRunExtensions.so ADDED
Binary file (423 kB). View file
 
genie_bundle_llama_v3_2_3b_gen_2/libQnnDspV66CalculatorStub.so ADDED
Binary file (7.82 kB). View file
 
genie_bundle_llama_v3_2_3b_gen_2/libQnnDspV66Stub.so ADDED
Binary file (313 kB). View file
 
genie_bundle_llama_v3_2_3b_gen_2/libQnnGenAiTransformer.so ADDED
Binary file (842 kB). View file
 
genie_bundle_llama_v3_2_3b_gen_2/libQnnGenAiTransformerCpuOpPkg.so ADDED
Binary file (558 kB). View file
 
genie_bundle_llama_v3_2_3b_gen_2/libQnnGenAiTransformerModel.so ADDED
Binary file (235 kB). View file
 
genie_bundle_llama_v3_2_3b_gen_2/libQnnGpu.so ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:ba9dbc4ecf5ed94a0cf87098f1cd4bb3c6e7e68e5e17af2db0d15e3d710a414b
3
+ size 5781424
genie_bundle_llama_v3_2_3b_gen_2/libQnnGpuNetRunExtensions.so ADDED
Binary file (415 kB). View file
 
genie_bundle_llama_v3_2_3b_gen_2/libQnnGpuProfilingReader.so ADDED
Binary file (594 kB). View file
 
genie_bundle_llama_v3_2_3b_gen_2/libQnnHta.so ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:0ca8f73b651fd423bef0ee21e9f6d8bfb69fdd2f5a51a2d0e04f49880f605eb1
3
+ size 1004136
genie_bundle_llama_v3_2_3b_gen_2/libQnnHtaNetRunExtensions.so ADDED
Binary file (410 kB). View file
 
genie_bundle_llama_v3_2_3b_gen_2/libQnnHtp.so ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:d553d5322ecabcfeb25d2c8fa6c3399b13f1ae229087362a790f7f835deee054
3
+ size 1961144
genie_bundle_llama_v3_2_3b_gen_2/libQnnHtpNetRunExtensions.so ADDED
Binary file (668 kB). View file
 
genie_bundle_llama_v3_2_3b_gen_2/libQnnHtpOptraceProfilingReader.so ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:727644d7d8a4226ed5c4fc18e7f9ff1879109d8e95c298a17c3803c0858dcfc9
3
+ size 2584888
genie_bundle_llama_v3_2_3b_gen_2/libQnnHtpPrepare.so ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:024346964a7d8a09d3efcfdf0e5d3c90a11feb8285104b5bc90066548579176a
3
+ size 45683400
genie_bundle_llama_v3_2_3b_gen_2/libQnnHtpProfilingReader.so ADDED
Binary file (579 kB). View file
 
genie_bundle_llama_v3_2_3b_gen_2/libQnnHtpV68CalculatorStub.so ADDED
Binary file (7.89 kB). View file
 
genie_bundle_llama_v3_2_3b_gen_2/libQnnHtpV68Stub.so ADDED
Binary file (332 kB). View file
 
genie_bundle_llama_v3_2_3b_gen_2/libQnnHtpV69CalculatorStub.so ADDED
Binary file (7.89 kB). View file
 
genie_bundle_llama_v3_2_3b_gen_2/libQnnHtpV69Stub.so ADDED
Binary file (334 kB). View file
 
genie_bundle_llama_v3_2_3b_gen_2/libQnnHtpV73.so ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:00d9eb1292e3d85f176619d5890fec23c686499066b85c424f9d2ffcd51afff7
3
+ size 11479400
genie_bundle_llama_v3_2_3b_gen_2/libQnnHtpV73CalculatorStub.so ADDED
Binary file (8.43 kB). View file
 
genie_bundle_llama_v3_2_3b_gen_2/libQnnHtpV73QemuDriver.so ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:ca98c8bfb34760958b0c2d4e823ff71a53f4a9623e7f2592a26b4d40cd4aeb8b
3
+ size 8832384
genie_bundle_llama_v3_2_3b_gen_2/libQnnHtpV73Skel.so ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:6ad7acfd5f2a5dcbad19aa16bedb75f7e60a579de1db20e1f8c4656f5a7a6c7e
3
+ size 8579080
genie_bundle_llama_v3_2_3b_gen_2/libQnnHtpV73Stub.so ADDED
Binary file (455 kB). View file
 
genie_bundle_llama_v3_2_3b_gen_2/libQnnHtpV75CalculatorStub.so ADDED
Binary file (8.15 kB). View file
 
genie_bundle_llama_v3_2_3b_gen_2/libQnnHtpV75Stub.so ADDED
Binary file (455 kB). View file
 
genie_bundle_llama_v3_2_3b_gen_2/libQnnHtpV79CalculatorStub.so ADDED
Binary file (8.15 kB). View file
 
genie_bundle_llama_v3_2_3b_gen_2/libQnnHtpV79Stub.so ADDED
Binary file (455 kB). View file
 
genie_bundle_llama_v3_2_3b_gen_2/libQnnJsonProfilingReader.so ADDED
Binary file (602 kB). View file
 
genie_bundle_llama_v3_2_3b_gen_2/libQnnLpai.so ADDED
Binary file (595 kB). View file
 
genie_bundle_llama_v3_2_3b_gen_2/libQnnLpaiNetRunExtensions.so ADDED
Binary file (678 kB). View file
 
genie_bundle_llama_v3_2_3b_gen_2/libQnnLpaiV79Stub.so ADDED
Binary file (291 kB). View file
 
genie_bundle_llama_v3_2_3b_gen_2/libQnnModelDlc.so ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:a41ac76422f2f386a1f5038cb62cc564c322c22f283ca2fde230f40f86b5fecc
3
+ size 1931656
genie_bundle_llama_v3_2_3b_gen_2/libQnnNetRunDirectV79Stub.so ADDED
Binary file (365 kB). View file
 
genie_bundle_llama_v3_2_3b_gen_2/libQnnSaver.so ADDED
Binary file (761 kB). View file
 
genie_bundle_llama_v3_2_3b_gen_2/libQnnSystem.so ADDED
Binary file (272 kB). View file
 
genie_bundle_llama_v3_2_3b_gen_2/libQnnTFLiteDelegate.so ADDED
Binary file (993 kB). View file
 
genie_bundle_llama_v3_2_3b_gen_2/libSNPE.so ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:578cdb549732dc866302f351ecb1c86bad44ac2e55f41214916287f6eb350512
3
+ size 17776320
genie_bundle_llama_v3_2_3b_gen_2/libSnpeDspV66Stub.so ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:91e57e5398e098c06d6a6b3421073fb7575a018e5530afe96244026793dc5649
3
+ size 1461424