franco-rojas commited on
Commit
a6082c6
1 Parent(s): 80ea7ba

Training in progress, step 500

Browse files
Files changed (3) hide show
  1. config.json +38 -0
  2. pytorch_model.bin +3 -0
  3. training_args.bin +3 -0
config.json ADDED
@@ -0,0 +1,38 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "_name_or_path": "tiiuae/falcon-rw-1b",
3
+ "alibi": true,
4
+ "apply_residual_connection_post_layernorm": false,
5
+ "architectures": [
6
+ "FalconForCausalLM"
7
+ ],
8
+ "attention_dropout": 0.0,
9
+ "auto_map": {
10
+ "AutoConfig": "tiiuae/falcon-rw-1b--configuration_falcon.FalconConfig",
11
+ "AutoModel": "tiiuae/falcon-rw-1b--modeling_falcon.FalconModel",
12
+ "AutoModelForCausalLM": "tiiuae/falcon-rw-1b--modeling_falcon.FalconForCausalLM",
13
+ "AutoModelForQuestionAnswering": "tiiuae/falcon-rw-1b--modeling_falcon.FalconForQuestionAnswering",
14
+ "AutoModelForSequenceClassification": "tiiuae/falcon-rw-1b--modeling_falcon.FalconForSequenceClassification",
15
+ "AutoModelForTokenClassification": "tiiuae/falcon-rw-1b--modeling_falcon.FalconForTokenClassification"
16
+ },
17
+ "bias": true,
18
+ "bos_token_id": 1,
19
+ "eos_token_id": 2,
20
+ "hidden_dropout": 0.0,
21
+ "hidden_size": 2048,
22
+ "initializer_range": 0.02,
23
+ "layer_norm_epsilon": 1e-05,
24
+ "max_position_embeddings": 2048,
25
+ "model_type": "falcon",
26
+ "multi_query": false,
27
+ "new_decoder_architecture": false,
28
+ "num_attention_heads": 32,
29
+ "num_hidden_layers": 24,
30
+ "num_kv_heads": 32,
31
+ "parallel_attn": false,
32
+ "rope_scaling": null,
33
+ "rope_theta": 10000.0,
34
+ "torch_dtype": "float32",
35
+ "transformers_version": "4.33.3",
36
+ "use_cache": true,
37
+ "vocab_size": 50304
38
+ }
pytorch_model.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:96dffbe24bafa212a85e87b3b5fdbf26dea33e4fad1581d13762ab4fae6a8fd4
3
+ size 5246599989
training_args.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:9fa240378d30b9f08125a63e1097a02afa5ef267f3e8b8f4ae11a123d754ce36
3
+ size 4091