Training checkpoint 114000 - 1.2B tokens

README.md

@@ -18,10 +18,10 @@ pipeline_tag: text-generation
 
 ## Current Training Status
 
-- **Training Step**: 112,000
-- **Tokens Processed**: 1.15B tokens
-- **Current Loss**: 4.5058
-- **Spike Rate**: 0.0504
+- **Training Step**: 114,000
+- **Tokens Processed**: 1.17B tokens
+- **Current Loss**: 4.6157
+- **Spike Rate**: 0.0503
 - **Learning Rate**: 1.98e-04
 
 ## Model Architecture
@@ -54,7 +54,7 @@ This represents ongoing training of the first large-scale spiking neural network
 from huggingface_hub import hf_hub_download
 checkpoint = hf_hub_download(
     repo_id="rootxhacker/piking-llm-5b-3epochs-exp",
-    filename="checkpoint_112000.pt"
+    filename="checkpoint_114000.pt"
 )
 
 # Load with custom spiking model code
@@ -65,4 +65,4 @@ checkpoint = hf_hub_download(
 
 **🔬 This is live research in progress! Check back for updates as training continues.**
 
-**Training Progress**: 7.6% complete towards 15B tokens
+**Training Progress**: 7.8% complete towards 15B tokens

checkpoint_114000.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:fe85299985f00bcf63d8f1f37d86ab55271325235d2fe523ce603fe7456f3ac5
+size 999026666

config.json

@@ -4,11 +4,11 @@
   "hidden_size": 768,
   "num_layers": 12,
   "max_seq_length": 1024,
-  "training_step": 112000,
-  "tokens_processed": 1146880000,
-  "loss": 4.505839907695941,
-  "spike_rate": 0.05036780015042009,
-  "learning_rate": 0.000198367815094099,
-  "epoch": 0.229376,
-  "progress_percent": 7.645886240135441
+  "training_step": 114000,
+  "tokens_processed": 1167360000,
+  "loss": 4.615654933585256,
+  "spike_rate": 0.05030253979607746,
+  "learning_rate": 0.00019828814962311595,
+  "epoch": 0.233472,
+  "progress_percent": 7.782419922995003
 }