denizyuret-shallowai commited on
Commit
cb95e0b
1 Parent(s): 9a64e09

Upload model

Browse files
Files changed (3) hide show
  1. config.json +1 -1
  2. modeling_custom.py +1 -1
  3. pytorch_model.bin +2 -2
config.json CHANGED
@@ -24,7 +24,7 @@
24
  "rotary_emb_base": 10000,
25
  "rotary_pct": 0.25,
26
  "tie_word_embeddings": false,
27
- "torch_dtype": "float32",
28
  "transformers_version": "4.31.0",
29
  "use_cache": true,
30
  "use_parallel_residual": true,
 
24
  "rotary_emb_base": 10000,
25
  "rotary_pct": 0.25,
26
  "tie_word_embeddings": false,
27
+ "torch_dtype": "float16",
28
  "transformers_version": "4.31.0",
29
  "use_cache": true,
30
  "use_parallel_residual": true,
modeling_custom.py CHANGED
@@ -18,7 +18,7 @@ class CustomModel(GPTNeoXForCausalLM):
18
  @classmethod
19
  def copy_from_neox(cls, *args, **kwargs):
20
  m0 = GPTNeoXForCausalLM.from_pretrained(*args, **kwargs)
21
- m1 = cls(m0.config)
22
  m1.load_state_dict(m0.state_dict())
23
  return m1
24
 
 
18
  @classmethod
19
  def copy_from_neox(cls, *args, **kwargs):
20
  m0 = GPTNeoXForCausalLM.from_pretrained(*args, **kwargs)
21
+ m1 = cls(m0.config).to(dtype=m0.dtype, device=m0.device)
22
  m1.load_state_dict(m0.state_dict())
23
  return m1
24
 
pytorch_model.bin CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:3e4128420302e58c85fc79e4553c320c270cab3169bc72310bdf2b5b75032a6d
3
- size 281732711
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:781eaeb60eaee1da015e90d41470942b4597682f8ead7b94aa29676139808cd7
3
+ size 140879463