hranjan043 commited on
Commit
3435e6e
·
verified ·
1 Parent(s): 74b93fd

Upload folder using huggingface_hub

Browse files
Files changed (2) hide show
  1. README.md +30 -13
  2. simbot.safetensors +3 -0
README.md CHANGED
@@ -13,32 +13,47 @@ framework: pytorch
13
 
14
  # SimBot GPT (Level 1)
15
 
16
- This is a **from-scratch GPT-style language model** trained using PyTorch.
 
17
 
18
- ### Training
19
- - Architecture: Decoder-only Transformer
20
- - Objective: Causal Language Modeling
21
- - Dataset: Simdega / domain-specific text
22
- - Purpose: Learning LLM internals (not instruction-tuned)
 
 
 
 
 
 
 
23
 
24
- ### Files
25
- - `simbot.pt` — model weights
 
26
  - `tokenizer.json` — BPE tokenizer
27
- - `model/simbot.py` — model architecture
 
 
 
28
 
29
- ### Usage (example)
30
 
31
  ```python
32
- import torch
 
33
  from tokenizers import Tokenizer
34
  from model.simbot import SIMGPT
35
- import json
36
 
 
37
  tokenizer = Tokenizer.from_file("tokenizer.json")
38
 
 
39
  with open("config.json") as f:
40
  cfg = json.load(f)
41
 
 
42
  model = SIMGPT(
43
  vocab_size=cfg["vocab_size"],
44
  block_size=cfg["block_size"],
@@ -47,5 +62,7 @@ model = SIMGPT(
47
  d_model=cfg["d_model"]
48
  )
49
 
50
- model.load_state_dict(torch.load("gpt.pt", map_location="cpu"))
 
 
51
  model.eval()
 
13
 
14
  # SimBot GPT (Level 1)
15
 
16
+ SimBot GPT is a **from-scratch GPT-style language model** implemented in **PyTorch**.
17
+ This project is focused on **learning LLM internals**, not on instruction tuning or production use.
18
 
19
+ ---
20
+
21
+ ## Model Overview
22
+
23
+ - **Architecture:** Decoder-only Transformer (GPT-like)
24
+ - **Training Objective:** Causal Language Modeling
25
+ - **Dataset:** Domain-specific text (Simdega / regional data)
26
+ - **Purpose:** Educational (understanding how LLMs work internally)
27
+
28
+ ⚠️ This is a **base language model**, not instruction-tuned and not grounded with RAG.
29
+
30
+ ---
31
 
32
+ ## Repository Contents
33
+
34
+ - `simbot.safetensors` — model weights (safe & HF-recommended format)
35
  - `tokenizer.json` — BPE tokenizer
36
+ - `config.json` — model hyperparameters
37
+ - `model/simbot.py` — model architecture (PyTorch)
38
+
39
+ ---
40
 
41
+ ## Usage Example
42
 
43
  ```python
44
+ import json
45
+ from safetensors.torch import load_file
46
  from tokenizers import Tokenizer
47
  from model.simbot import SIMGPT
 
48
 
49
+ # Load tokenizer
50
  tokenizer = Tokenizer.from_file("tokenizer.json")
51
 
52
+ # Load config
53
  with open("config.json") as f:
54
  cfg = json.load(f)
55
 
56
+ # Build model
57
  model = SIMGPT(
58
  vocab_size=cfg["vocab_size"],
59
  block_size=cfg["block_size"],
 
62
  d_model=cfg["d_model"]
63
  )
64
 
65
+ # Load weights
66
+ state_dict = load_file("simbot.safetensors")
67
+ model.load_state_dict(state_dict)
68
  model.eval()
simbot.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:025e9428d136cc01078962849d732c1ee63dd73899f995246893ca58e0ab4b97
3
+ size 48078808