From a1b59b8792535d4728580c65392957057376a7b7 Mon Sep 17 00:00:00 2001
From: dohe0342 <kimdohe1070@gmail.com>
Date: Mon, 26 Dec 2022 13:58:38 +0900
Subject: [PATCH] from local

---
 .../.train.py.swp                             | Bin 94208 -> 94208 bytes
 .../train.py                                  |  60 +-----------------
 2 files changed, 2 insertions(+), 58 deletions(-)

diff --git a/egs/librispeech/ASR/pruned_transducer_stateless_d2v_v2/.train.py.swp b/egs/librispeech/ASR/pruned_transducer_stateless_d2v_v2/.train.py.swp
index a1ce21a4c02dc4f8316ed0280000a7df75c41ebd..a736ebcc59df3a1f3c88b6ef48e834f7f1e6a177 100644
GIT binary patch
delta 770
zcmb8rO(+Cm7zgm@#n=x!>`pr?lt{v5%$9s4Id57y_$bj9t)o^sptLKFa(UId?1j@F
z9JGg3a=>21iQ*!ba#4hwe8m67JFVhi>NoSuy#M!kW+s+ZVp&BWG5Z6-sy4$U9B!i1
zhU94PhF+mDb>6R~{Te&&P*W;Ve(_g8jeB}nP4peN&#(py&;=rr2XBbKg4mXX85o3i
zsDKw2(Ic$G1VmsAba*NvI)OttfE5UX7xtV)yRhk`C@rBFhc0LWHx$B|gJ=s@U=D_1
z2wK4l*LI>ENWe7ofgkQ{7zhbihFKVcFmyu;G{a4y_){sRU{T>04|YCvuCfr1oa<lb
z_-!@o<c<3Zugo~|LSeBYE}o1p=kC(b#kvTWiG(PMWnsvo<Ud9EFJTN7oj?10PyaE=
zCHLt3((uVQ`ZHdX8!we{@wN82Iluz4sF%4)w6$^$`P7>Fxs%@;F4?M1Moet9{*X1+
mDx5NBJt@zfB>LK9&nu-02gSE*S&$#C`grzPOHYQ_qWc4VVB_lm

delta 375
zcmXBQKS%;`7{>AU2c4;xWrx;KkrYKi$Qd*^)Ra(2+Ehl6U{I4o(ke<plT9}q4M8n|
z(Gn>)87*#sK}(zVhoCW{rNR1EFMQw~o_BbkVoNKwv|z?uSzHQ5!=^BPk-N$AX81BV
zV5r!B#Bd^px}WuZY9eEu|ACm@yR1eF@g5NqFWn-~xJ4Be9OD3aBr%0IpGY02$YTvN
zaN(&-<Qg>`VGn7{ApkF4b&&=xP|~F!8I}}Q;X>8RA(BX79^=qZ@razGj4UQFjBmF{
z6E`@)4z{s{Aq=9<-)G3;uY=#B<jT|Wpjpj#w;fF>-FgfM?9jSeu-YGec6LJ@I*k!k
Wb6N3r&Yn)FeyjCmIJ;?8^8WzB8%FE^

diff --git a/egs/librispeech/ASR/pruned_transducer_stateless_d2v_v2/train.py b/egs/librispeech/ASR/pruned_transducer_stateless_d2v_v2/train.py
index 363edca92..519c78e63 100755
--- a/egs/librispeech/ASR/pruned_transducer_stateless_d2v_v2/train.py
+++ b/egs/librispeech/ASR/pruned_transducer_stateless_d2v_v2/train.py
@@ -1528,64 +1528,8 @@ def run_adapter(rank, world_size, args, wb=None):
             clipping_scale=None,
             parameters_names=[adapter_names],
         )
-
-    if params.multi_optim:
-        logging.info("Using seperate optimizers over encoder, decoder ...")
-
-        enc_param = []
-        enc_names = []
-
-        dec_names = []
-        dec_param = []
-        
-        for n, p in model.named_parameters():
-            name = n.split('.')[1]
-            if name == 'encoder' and 'feature_extractor' not in n:
-                enc_names.append(n)
-                enc_param.append(p)
-            elif 'ctc_output' in n:
-                enc_names.append(n)
-                enc_param.append(p)
-            elif 'feature_extractor' not in n:
-                dec_names.append(n)
-                dec_param.append(p)
-
-        optimizer_enc = ScaledAdam(
-            enc_param,
-            lr=params.peak_enc_lr,
-            clipping_scale=None,
-            parameters_names=[enc_names],
-        )
-        optimizer_dec = ScaledAdam(
-            dec_param,
-            lr=params.peak_dec_lr,
-            clipping_scale=5.0,
-            parameters_names=[dec_names],
-        )
-
-        scheduler_enc = Eden(optimizer_enc, params.lr_batches, params.lr_epochs)
-        scheduler_dec = Eden(optimizer_dec, params.lr_batches, params.lr_epochs)
-        optimizer = [optimizer_enc, optimizer_dec]
-        scheduler = [scheduler_enc, scheduler_dec]
-
-    else:
-        parameters_names = []
-        parameters_names.append(
-            [name_param_pair[0] for name_param_pair in model.named_parameters()]
-        )
-
-        logging.info(f"len name = {len(parameters_names)}")
-        logging.info(f"len param = {len(list(model.parameters()))}")
-        
-        optimizer = ScaledAdam(
-            model.parameters(),
-            lr=params.base_lr,
-            clipping_scale=2.0,
-            parameters_names=parameters_names,
-        )
-
-        scheduler = Eden(optimizer, params.lr_batches, params.lr_epochs)
-
+    scheduler_enc = Eden(optimizer_enc, params.lr_batches, params.lr_epochs)
+    
     if checkpoints and ("optimizer" in checkpoints or "optimizer_enc" in checkpoints):
         if params.multi_optim:
             logging.info("Loading optimizer state dict")