Training in progress, step 3

adapter_config.json

@@ -16,12 +16,12 @@
   "megatron_core": "megatron.core",
   "modules_to_save": null,
   "peft_type": "LORA",
-  "r": 16,
+  "r": 32,
   "rank_pattern": {},
   "revision": null,
   "target_modules": [
-    "v_proj",
-    "q_proj"
+    "q_proj",
+    "v_proj"
   ],
   "task_type": "CAUSAL_LM",
   "use_dora": false,

adapter_model.safetensors

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:930c2c8531a25aee78949e34103db4f55f9c31db7ddc3a11246e072d3bc24a55
-size 27280152
+oid sha256:37d3d38b9a505fbc963fda4bc2827b8d97df5f815f3adbf6052296d38d476548
+size 54543184

config.json

@@ -1,24 +1,35 @@
 {
-    "dataset_name": "data/packaged_pretrain_dataset.parquet",
-    "num_proc": 1,
-    "max_seq_length": 32,
-    "seed": 0,
-    "optim": "adamw_torch",
-    "max_steps": 10000,
-    "per_device_train_batch_size": 2,
-    "learning_rate": 5e-05,
-    "weight_decay": 0,
-    "warmup_steps": 10,
-    "lr_scheduler_type": "linear",
-    "gradient_checkpointing": true,
-    "dataloader_num_workers": 2,
-    "bf16": true,
-    "gradient_accumulation_steps": 1,
-    "logging_steps": 3,
-    "report_to": [],
-    "save_strategy": "steps",
-    "save_steps": 3,
-    "save_total_limit": 1,
-    "push_to_hub": true,
-    "hub_model_id": "hienbm/psychology-llama3.1-8B"
-}
+  "_name_or_path": "meta-llama/Meta-Llama-3.1-8B",
+  "architectures": [
+    "LlamaForCausalLM"
+  ],
+  "attention_bias": false,
+  "attention_dropout": 0.0,
+  "bos_token_id": 128000,
+  "eos_token_id": 128001,
+  "hidden_act": "silu",
+  "hidden_size": 1024,
+  "initializer_range": 0.02,
+  "intermediate_size": 4096,
+  "max_position_embeddings": 131072,
+  "mlp_bias": false,
+  "model_type": "llama",
+  "num_attention_heads": 32,
+  "num_hidden_layers": 16,
+  "num_key_value_heads": 8,
+  "pretraining_tp": 1,
+  "rms_norm_eps": 1e-05,
+  "rope_scaling": {
+    "factor": 8.0,
+    "high_freq_factor": 4.0,
+    "low_freq_factor": 1.0,
+    "original_max_position_embeddings": 8192,
+    "rope_type": "llama3"
+  },
+  "rope_theta": 500000.0,
+  "tie_word_embeddings": false,
+  "torch_dtype": "float32",
+  "transformers_version": "4.44.0.dev0",
+  "use_cache": false,
+  "vocab_size": 128256
+}

training_args.bin

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:0bb5ecac405a641762d526405359377cb588451eb0138aa06cb6661dd108a933
+oid sha256:ac8a9bdf4b4f3d949455dea461e328a1830a46efff04572a20bcb39b4de997a8
 size 5304