dacorvo HF Staff commited on
Commit
0f6c357
·
verified ·
1 Parent(s): 3f1cc1e

Synchronizing local compiler cache.

Browse files
.gitattributes CHANGED
@@ -17786,3 +17786,6 @@ neuronxcc-2.26.6360.0+6f180f47/MODULE_f0727109886547b51ba5+12dd9119/model.neff f
17786
  neuronxcc-2.26.6360.0+6f180f47/MODULE_5b01ff4df14f87503b4e+26701361/model.neff filter=lfs diff=lfs merge=lfs -text
17787
  neuronxcc-2.26.6360.0+6f180f47/MODULE_63a739e472c9747f77fd+8f3c2f7c/model.neff filter=lfs diff=lfs merge=lfs -text
17788
  neuronxcc-2.26.6360.0+6f180f47/MODULE_63a739e472c9747f77fd+8f3c2f7c/wrapped_neff.hlo filter=lfs diff=lfs merge=lfs -text
 
 
 
 
17786
  neuronxcc-2.26.6360.0+6f180f47/MODULE_5b01ff4df14f87503b4e+26701361/model.neff filter=lfs diff=lfs merge=lfs -text
17787
  neuronxcc-2.26.6360.0+6f180f47/MODULE_63a739e472c9747f77fd+8f3c2f7c/model.neff filter=lfs diff=lfs merge=lfs -text
17788
  neuronxcc-2.26.6360.0+6f180f47/MODULE_63a739e472c9747f77fd+8f3c2f7c/wrapped_neff.hlo filter=lfs diff=lfs merge=lfs -text
17789
+ neuronxcc-2.26.6360.0+6f180f47/MODULE_0894513edb43964963c4+26701361/model.neff filter=lfs diff=lfs merge=lfs -text
17790
+ neuronxcc-2.26.6360.0+6f180f47/MODULE_88dfbc29988e4b42dc0f+8f3c2f7c/model.neff filter=lfs diff=lfs merge=lfs -text
17791
+ neuronxcc-2.26.6360.0+6f180f47/MODULE_88dfbc29988e4b42dc0f+8f3c2f7c/wrapped_neff.hlo filter=lfs diff=lfs merge=lfs -text
neuronxcc-2.26.6360.0+6f180f47/0_REGISTRY/0.4.7.dev0/2aae7cc49f53a364c1bfabafaecff51826622f8f3ade9b884d1f5aa9b6107830/da97150a8d3791f34120.json ADDED
@@ -0,0 +1,59 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "_entry_class": "SingleModelCacheEntry",
3
+ "_model_id": "mistralai/Mixtral-8x7B-Instruct-v0.1",
4
+ "_task": "text-generation",
5
+ "architectures": [
6
+ "MixtralForCausalLM"
7
+ ],
8
+ "attention_dropout": 0.0,
9
+ "dtype": "bfloat16",
10
+ "head_dim": 128,
11
+ "hidden_act": "silu",
12
+ "hidden_size": 4096,
13
+ "initializer_range": 0.02,
14
+ "intermediate_size": 14336,
15
+ "max_position_embeddings": 32768,
16
+ "model_type": "mixtral",
17
+ "neuron": {
18
+ "_serialized_key": "NxDNeuronConfig",
19
+ "batch_size": 4,
20
+ "capacity_factor": null,
21
+ "checkpoint_id": "mistralai/Mixtral-8x7B-Instruct-v0.1",
22
+ "checkpoint_revision": "eba92302a2861cdc0098cc54bc9f17cb2c47eb61",
23
+ "continuous_batching": false,
24
+ "ep_degree": 1,
25
+ "fused_qkv": false,
26
+ "glu_mlp": true,
27
+ "local_ranks_size": 32,
28
+ "max_batch_size": 4,
29
+ "max_context_length": 4096,
30
+ "max_topk": 256,
31
+ "n_active_tokens": 4096,
32
+ "neuronxcc_version": "2.26.6360.0+6f180f47",
33
+ "on_device_sampling": false,
34
+ "optimum_neuron_version": "0.4.7.dev0",
35
+ "output_logits": false,
36
+ "pp_degree": 1,
37
+ "prefill_chunk_size": 1024,
38
+ "sequence_length": 4096,
39
+ "speculation_length": 0,
40
+ "start_rank_id": 0,
41
+ "target": "trn1",
42
+ "torch_dtype": "bfloat16",
43
+ "tp_degree": 32
44
+ },
45
+ "num_attention_heads": 32,
46
+ "num_experts_per_tok": 2,
47
+ "num_hidden_layers": 32,
48
+ "num_key_value_heads": 8,
49
+ "num_local_experts": 8,
50
+ "output_router_logits": false,
51
+ "rms_norm_eps": 1e-05,
52
+ "rope_theta": 1000000.0,
53
+ "router_aux_loss_coef": 0.02,
54
+ "router_jitter_noise": 0.0,
55
+ "sliding_window": null,
56
+ "tie_word_embeddings": false,
57
+ "use_cache": true,
58
+ "vocab_size": 32000
59
+ }
neuronxcc-2.26.6360.0+6f180f47/MODULE_0894513edb43964963c4+26701361/compile_flags.json ADDED
@@ -0,0 +1 @@
 
 
1
+ ["--target=inf2", "--enable-saturate-infinity", "--enable-mixed-precision-accumulation", "--model-type", "transformer", "-O1", "--tensorizer-options=--enable-ccop-compute-overlap --cc-pipeline-tiling-factor=2", "--auto-cast=none", "--internal-enable-dge-levels", "vector_dynamic_offsets", "--verbose=35", "--logfile=/tmp/nxd_model/chunked_prefill/_tp0_bk0/log-neuron-cc.txt"]
neuronxcc-2.26.6360.0+6f180f47/MODULE_0894513edb43964963c4+26701361/model.done ADDED
File without changes
neuronxcc-2.26.6360.0+6f180f47/MODULE_0894513edb43964963c4+26701361/model.hlo_module.pb ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:d5102d3b71d883e0df840f9995772fffc1f4d402c6e61d391f35ace7ae0ac13d
3
+ size 59097341
neuronxcc-2.26.6360.0+6f180f47/MODULE_0894513edb43964963c4+26701361/model.neff ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:d7c62cb07b3a71dd82da47b201e39e2496d49c464e1ebadc6e4162bf50335adb
3
+ size 6093824
neuronxcc-2.26.6360.0+6f180f47/MODULE_88dfbc29988e4b42dc0f+8f3c2f7c/compile_flags.json ADDED
@@ -0,0 +1 @@
 
 
1
+ ["--target=inf2", "--enable-saturate-infinity", "--enable-mixed-precision-accumulation", "--model-type", "transformer", "-O1", "--tensorizer-options=--enable-ccop-compute-overlap --cc-pipeline-tiling-factor=2", "--auto-cast=none", "--internal-enable-dge-levels", "vector_dynamic_offsets", "--verbose=35", "--logfile=/tmp/nxd_model/token_generation/_tp0_bk0/log-neuron-cc.txt", "--enable-internal-neff-wrapper"]
neuronxcc-2.26.6360.0+6f180f47/MODULE_88dfbc29988e4b42dc0f+8f3c2f7c/model.done ADDED
File without changes
neuronxcc-2.26.6360.0+6f180f47/MODULE_88dfbc29988e4b42dc0f+8f3c2f7c/model.hlo_module.pb ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:909fdd7a3dd6397a73e279674f3370951d5fa8a825764fdd39a206db69546563
3
+ size 5785389
neuronxcc-2.26.6360.0+6f180f47/MODULE_88dfbc29988e4b42dc0f+8f3c2f7c/model.neff ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:363ee460e69ab5f05ba66bd7572365172b2b2e1e81b361a98b3024f5d645cdfb
3
+ size 4967424
neuronxcc-2.26.6360.0+6f180f47/MODULE_88dfbc29988e4b42dc0f+8f3c2f7c/wrapped_neff.hlo ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:628dab90decdb77f54c06aa2c29c2459c5b92e9134aa52e86277e12292cbc2f5
3
+ size 5138152