Synchronizing local compiler cache.
Browse files
neuronxcc-2.17.194.0+d312836f/0_REGISTRY/0.2.0.dev3/inference/llama/unsloth/Llama-3.2-1B-Instruct/09d415999f129eb1a95e.json
ADDED
|
@@ -0,0 +1,56 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"_entry_class": "SingleModelCacheEntry",
|
| 3 |
+
"_model_id": "unsloth/Llama-3.2-1B-Instruct",
|
| 4 |
+
"_task": "text-generation",
|
| 5 |
+
"architectures": [
|
| 6 |
+
"LlamaForCausalLM"
|
| 7 |
+
],
|
| 8 |
+
"attention_bias": false,
|
| 9 |
+
"attention_dropout": 0.0,
|
| 10 |
+
"head_dim": 64,
|
| 11 |
+
"hidden_act": "silu",
|
| 12 |
+
"hidden_size": 2048,
|
| 13 |
+
"initializer_range": 0.02,
|
| 14 |
+
"intermediate_size": 8192,
|
| 15 |
+
"max_position_embeddings": 131072,
|
| 16 |
+
"mlp_bias": false,
|
| 17 |
+
"model_type": "llama",
|
| 18 |
+
"neuron": {
|
| 19 |
+
"all_reduce_dtype": null,
|
| 20 |
+
"allow_flash_attention": true,
|
| 21 |
+
"attention_layout": "BSH",
|
| 22 |
+
"attn_output_transposed": false,
|
| 23 |
+
"auto_cast_type": "fp16",
|
| 24 |
+
"backend": "hlo",
|
| 25 |
+
"batch_size": 4,
|
| 26 |
+
"checkpoint_id": "unsloth/Llama-3.2-1B-Instruct",
|
| 27 |
+
"checkpoint_revision": "9b58d4a36161a1e49ecf0a69d20b2736fef8e438",
|
| 28 |
+
"collectives_layout": "HSB",
|
| 29 |
+
"continuous_batching": true,
|
| 30 |
+
"fuse_qkv": true,
|
| 31 |
+
"group_query_attention": "shard-over-heads",
|
| 32 |
+
"log_softmax_scores": false,
|
| 33 |
+
"neuronxcc_version": "2.17.194.0+d312836f",
|
| 34 |
+
"optimum_neuron_version": "0.2.0.dev3",
|
| 35 |
+
"output_all_logits": false,
|
| 36 |
+
"sequence_length": 4096,
|
| 37 |
+
"tp_degree": 2
|
| 38 |
+
},
|
| 39 |
+
"num_attention_heads": 32,
|
| 40 |
+
"num_hidden_layers": 16,
|
| 41 |
+
"num_key_value_heads": 8,
|
| 42 |
+
"pretraining_tp": 1,
|
| 43 |
+
"rms_norm_eps": 1e-05,
|
| 44 |
+
"rope_scaling": {
|
| 45 |
+
"factor": 32.0,
|
| 46 |
+
"high_freq_factor": 4.0,
|
| 47 |
+
"low_freq_factor": 1.0,
|
| 48 |
+
"original_max_position_embeddings": 8192,
|
| 49 |
+
"rope_type": "llama3"
|
| 50 |
+
},
|
| 51 |
+
"rope_theta": 500000.0,
|
| 52 |
+
"tie_word_embeddings": true,
|
| 53 |
+
"unsloth_fixed": true,
|
| 54 |
+
"use_cache": true,
|
| 55 |
+
"vocab_size": 128256
|
| 56 |
+
}
|
neuronxcc-2.17.194.0+d312836f/MODULE_7a4e910c3dbc7ccf8eb3+613edded/model.neff
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 29072384
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:98bf65cdce6d9e04670d52b876bcbf1ef63801f425a23366708e18485a789902
|
| 3 |
size 29072384
|
neuronxcc-2.17.194.0+d312836f/MODULE_f91cd98a64f373af274d+613edded/model.neff
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 2448384
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:f8ffe7e9a80e9bb621d91d69c3ce56f8841902928b64fab653f11cc9762806af
|
| 3 |
size 2448384
|