Update instruction

.gitignore

@@ -1,2 +0,0 @@
-checkpoints/*
-!checkpoints/README.md

README.md

@@ -2,7 +2,7 @@
 
 This repository contains JAX example code for loading and running the Grok-1 open-weights model.
 
-Make sure to download the checkpoint and place the `ckpt-0` directory in `checkpoints` - see [Downloading the weights](#downloading-the-weights)
+Make sure to download the checkpoint and place `ckpt-0` directory in `checkpoint` - see [Downloading the weights](Downloading-the-weights)
 
 Then, run
 
@@ -18,31 +18,14 @@ The script loads the checkpoint and samples from the model on a test input.
 Due to the large size of the model (314B parameters), a machine with enough GPU memory is required to test the model with the example code.
 The implementation of the MoE layer in this repository is not efficient. The implementation was chosen to avoid the need for custom kernels to validate the correctness of the model.
 
-# Model Specifications
-
-Grok-1 is currently designed with the following specifications:
-
-- **Parameters:** 314B
-- **Architecture:** Mixture of 8 Experts (MoE)
-- **Experts Utilization:** 2 experts used per token
-- **Layers:** 64
-- **Attention Heads:** 48 for queries, 8 for keys/values
-- **Embedding Size:** 6,144
-- **Tokenization:** SentencePiece tokenizer with 131,072 tokens
-- **Additional Features:**
-  - Rotary embeddings (RoPE)
-  - Supports activation sharding and 8-bit quantization
-- **Maximum Sequence Length (context):** 8,192 tokens
-
 # Downloading the weights
 
 You can download the weights using a torrent client and this magnet link:
-
 ```
 magnet:?xt=urn:btih:5f96d43576e3d386c9ba65b883210a393b68210e&tr=https%3A%2F%2Facademictorrents.com%2Fannounce.php&tr=udp%3A%2F%2Ftracker.coppersurfer.tk%3A6969&tr=udp%3A%2F%2Ftracker.opentrackr.org%3A1337%2Fannounce
 ```
 
-or directly using [HuggingFace 🤗 Hub](https://huggingface.co/xai-org/grok-1):
+or directly using HuggingFace:
 ```
 git clone https://github.com/xai-org/grok-1.git && cd grok-1
 pip install huggingface_hub[hf_transfer]

requirements.txt

@@ -1,4 +1,4 @@
 dm_haiku==0.0.12
-jax[cuda12-pip]==0.4.25 -f https://storage.googleapis.com/jax-releases/jax_cuda_releases.html
+jax[cuda12_pip]==0.4.25 -f https://storage.googleapis.com/jax-releases/jax_cuda_releases.html
 numpy==1.26.4
 sentencepiece==0.2.0

run.py

@@ -13,86 +13,60 @@
 # limitations under the License.
 
 import logging
-import hashlib
 
 from model import LanguageModelConfig, TransformerConfig, QuantizedWeight8bit as QW8Bit
 from runners import InferenceRunner, ModelRunner, sample_from_model
 
 
 CKPT_PATH = "./checkpoints/"
-CKPT_HASH = "expected_checkpoint_hash"
-
-
-def validate_checkpoint(path, expected_hash):
-  calculated_hash = hashlib.sha256(open(path, 'rb').read()).hexdigest()
-  if calculated_hash != expected_hash:
-    raise ValueError("Invalid checkpoint file!")
 
 
 def main():
-  # Validate checkpoint integrity
-  validate_checkpoint(CKPT_PATH, CKPT_HASH)
+    grok_1_model = LanguageModelConfig(
+        vocab_size=128 * 1024,
+        pad_token=0,
+        eos_token=2,
+        sequence_len=8192,
+        embedding_init_scale=1.0,
+        output_multiplier_scale=0.5773502691896257,
+        embedding_multiplier_scale=78.38367176906169,
+        model=TransformerConfig(
+            emb_size=48 * 128,
+            widening_factor=8,
+            key_size=128,
+            num_q_heads=48,
+            num_kv_heads=8,
+            num_layers=64,
+            attn_output_multiplier=0.08838834764831845,
+            shard_activations=True,
+            # MoE.
+            num_experts=8,
+            num_selected_experts=2,
+            # Activation sharding.
+            data_axis="data",
+            model_axis="model",
+        ),
+    )
+    inference_runner = InferenceRunner(
+        pad_sizes=(1024,),
+        runner=ModelRunner(
+            model=grok_1_model,
+            bs_per_device=0.125,
+            checkpoint_path=CKPT_PATH,
+        ),
+        name="local",
+        load=CKPT_PATH,
+        tokenizer_path="./tokenizer.model",
+        local_mesh_config=(1, 8),
+        between_hosts_config=(1, 1),
+    )
+    inference_runner.initialize()
+    gen = inference_runner.run()
 
-  grok_1_model = LanguageModelConfig(
-    vocab_size=128 * 1024,
-    pad_token=0, 
-    eos_token=2,
-    sequence_len=8192,
-    embedding_init_scale=1.0,
-    output_multiplier_scale=0.5773502691896257,
-    embedding_multiplier_scale=78.38367176906169,
-    model=TransformerConfig(
-      emb_size=48 * 128,
-      widening_factor=8,
-      key_size=128,
-      num_q_heads=48,
-      num_kv_heads=8,
-      num_layers=64,
-      attn_output_multiplier=0.08838834764831845,
-      shard_activations=True,
-      # MoE.
-      num_experts=8,
-      num_selected_experts=2,
-      # Activation sharding.
-      data_axis="data",
-      model_axis="model",
-    ),
-  )
+    inp = "The answer to life the universe and everything is of course"
+    print(f"Output for prompt: {inp}", sample_from_model(gen, inp, max_len=100, temperature=0.01))
 
-  inference_runner = InferenceRunner(
-    pad_sizes=(1024,),
-    runner=ModelRunner(
-      model=grok_1_model,
-      bs_per_device=0.125, 
-      checkpoint_path=CKPT_PATH,
-    # Limit inference rate
-    inference_runner.rate_limit = 100
-    ),
-    
-    name="local",
-    load=CKPT_PATH,
-    tokenizer_path="./tokenizer.model",
-    local_mesh_config=(1, 8),
-    between_hosts_config=(1, 1),
-  )
-  
-  inference_runner.initialize()
-  
-  gen = inference_runner.run()
-
-  inp = "The answer to life the universe and everything is of course"
-  print(f"Output for prompt: {inp}", sample_from_model(gen, inp, max_len=100, temperature=0.01))
-
-  # Add authentication
-  @app.route("/inference")
-  @auth.login_required
-  def inference():
-    ...
-  
-  gen = inference_runner.run()
-
-  # Rest of inference code
 
 if __name__ == "__main__":
-  logging.basicConfig(level=logging.INFO)
-  main()
+    logging.basicConfig(level=logging.INFO)
+    main()