We read every piece of feedback, and take your input very seriously.
To see all available qualifiers, see our documentation.
There was an error while loading. Please reload this page.
1 parent 2ed6465 commit 7ad51b2Copy full SHA for 7ad51b2
xtra_labs/llm_finetune/draft.py
@@ -0,0 +1,34 @@
1
+"""
2
+Drafting lab flow in script format using PyTorch
3
4
+
5
+# Part 1
6
7
+# Load pretrained LLM (medium size model)
8
9
+# Function for next token prediction
10
11
+# Call in loop for autoregressive inference
12
13
+# Evaluate model on benchmark dataset
14
15
+# Do the same for small and large models
16
17
+# Spider plot
18
19
20
+# Part 2
21
22
+# new LoRA linear layer class
23
24
+# new attention layer class
25
26
+# replace attention modules with new module
27
28
+# load chat dataset
29
30
+# train model
31
32
+# evaluate finetuned model on benchmark
33
34
+# add to spider plot
0 commit comments