Upload folder using huggingface_hub

saves/checkpoint/final_model.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:8a6420bc0e4f6f90da4b294c70472353230f45f19e203d263edbc58a63785acc
+size 72399768

saves/checkpoint/num_steps_90000.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:5e7ccc794185d9efa15fb1622aa883213f1a4a1bbca7055afb1b8fac968b4366
+size 217200386

saves/ffnn/FeedForward_Network

@@ -0,0 +1,21 @@
+digraph {
+	Input [label="Input
+[B,Seq_Len,Dim_Model]" shape=box]
+	FC1 [label="Linear
+(dim_model---->dim_ff)" shape=box]
+	Split [label="Split(x--->x1, x2)" shape=box]
+	SiLU [label="SiLU(x1)" shape=box]
+	Gate [label="SiLU(x1) * x2
+(SwiGLU)" shape=box]
+	FC2 [label="Linear
+(dim_ff---->dim_model)" shape=box]
+	Output [label="Output
+[B,Seq_len,Dim_model]" shape=box]
+	Input -> FC1
+	FC1 -> Split
+	Split -> SiLU
+	Split -> Gate
+	SiLU -> Gate
+	Gate -> FC2
+	FC2 -> Output
+}

saves/mha/multihead_attention_flow

@@ -0,0 +1,32 @@
+digraph {
+	x [label="Input x
+[B,Seq_len,Dim_Model]"]
+	qkv [label="QKV Projection
+[B,Seq_len,3*Dim_Model]"]
+	reshape [label="Reshape
+[3,B,H,Seq_len,Dim_k]"]
+	Q [label="Query
+[B,H,Seq_len,Dim_Key]"]
+	K [label="Key
+[B,H,Seq_len,Dim_Key]"]
+	V [label="Value
+[B,H,Seq_len,Dim_Key]"]
+	rope [label="Rotary Positional Embedding"]
+	attn [label="Flash Attention
+[B, H, Seq_len, Dim_k]"]
+	merge [label="Merge Heads
+[B, Seq_len, Dim_model]"]
+	out_proj [label="Output Linear
+[B, Seq_len, Dim_model]"]
+	x -> qkv
+	qkv -> reshape
+	reshape -> Q
+	reshape -> K
+	reshape -> V
+	Q -> rope
+	K -> rope
+	rope -> attn [label="Q, K"]
+	V -> attn [label=V]
+	attn -> merge
+	merge -> out_proj
+}

saves/smollm/smollm_visualization

@@ -0,0 +1,22 @@
+digraph {
+	size=15
+	Tokens [label="Input Tokens
+[B, Seq_len]" shape=box]
+	Embed [label="Token Embedding
+[B, Seq_len, Dim_model]" shape=ellipse]
+	Scale [label="Scale by sqrt(dim_model)" shape=ellipse]
+	Mask [label="Causal Mask
+[seq_len * seq_len]" shape=diamond]
+	Transformer_Stack [label="8 Transformer Blocks" shape=box3d]
+	Norm [label=RMSNorm shape=ellipse]
+	Output [label="Linear Projection
+[B, Seq_len, Vocab_size]" shape=box]
+	Logits [label=Logits shape=box]
+	Tokens -> Embed
+	Embed -> Scale
+	Scale -> Transformer_Stack
+	Mask -> Transformer_Stack [label="(Optional)" style=dashed]
+	Transformer_Stack -> Norm
+	Norm -> Output
+	Output -> Logits
+}

saves/text_samples_10000_44c46ef0860634d197e0.table.json

@@ -0,0 +1 @@
+{"columns": ["Step", "Prompt", "Generated"], "data": [[10000, "Wednesday, May 25, 2011\n\n", "Wednesday, May 25, 2011\n\nThe foliina Niemenkari, 2011 Muslims are two of the whole wise to small cadets, a job. The"]]}

saves/tokenizer/bpe_tokenizer_16k_n1000000.pkl

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:69be2d907b6d9dc0986fcb980188eb7fcb3b0a8e92bc925573ad7fb68d685fd9
+size 194466

saves/tokenizer/essential_web_500k_tokens.pkl

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:264cbebdca2392db70f693520d53203baf61f725ca1894c35f8e65bff976904c
+size 349894

saves/transformer/Transformer_Block

@@ -0,0 +1,22 @@
+digraph {
+	Input [label="Input
+[B,Seq_len,D_Model]" shape=box]
+	Norm1 [label=RMSNorm shape=ellipse]
+	MHA [label="Multi-Head Attention" shape=box]
+	Res1 [label="Residual Connection" shape=diamond]
+	Norm2 [label=RMSNorm shape=ellipse]
+	FFNN [label="Feed-Forward
+(SwiGLU)" shape=box]
+	Res2 [label="Residual Connection" shape=diamond]
+	Output [label="Output
+[B,Seq_len,Dim_model]" shape=box]
+	Input -> Norm1
+	Input -> Res1
+	Norm1 -> MHA
+	MHA -> Res1
+	Res1 -> Norm2
+	Res1 -> Res2
+	Norm2 -> FFNN
+	FFNN -> Res2
+	Res2 -> Output
+}